code
stringlengths
81
54k
code_codestyle
int64
0
721
style_context
stringlengths
91
41.9k
style_context_codestyle
int64
0
699
label
int64
0
1
import json import os import re import unicodedata from json.encoder import INFINITY from typing import Any, Dict, List, Optional, Tuple, Union import numpy as np import regex from ...tokenization_utils import AddedToken, PreTrainedTokenizer from ...tokenization_utils_base import BatchEncoding from ...utils import TensorType, is_flax_available, is_tf_available, is_torch_available, logging from ...utils.generic import _is_jax, _is_numpy _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "artists_file": "artists.json", "lyrics_file": "lyrics.json", "genres_file": "genres.json", } _lowerCAmelCase = { "artists_file": { "jukebox": "https://huggingface.co/ArthurZ/jukebox/blob/main/artists.json", }, "genres_file": { "jukebox": "https://huggingface.co/ArthurZ/jukebox/blob/main/genres.json", }, "lyrics_file": { "jukebox": "https://huggingface.co/ArthurZ/jukebox/blob/main/lyrics.json", }, } _lowerCAmelCase = { "jukebox": 512, } class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = VOCAB_FILES_NAMES UpperCAmelCase = PRETRAINED_VOCAB_FILES_MAP UpperCAmelCase = PRETRAINED_LYRIC_TOKENS_SIZES UpperCAmelCase = ["input_ids", "attention_mask"] def __init__( self : int , _A : Optional[int] , _A : str , _A : Union[str, Any] , _A : List[str]=["v3", "v2", "v2"] , _A : int=512 , _A : List[Any]=5 , _A : Tuple="<|endoftext|>" , **_A : Optional[int] , ): _UpperCamelCase = AddedToken(_A , lstrip=_A , rstrip=_A ) if isinstance(_A , _A ) else unk_token super().__init__( unk_token=_A , n_genres=_A , version=_A , max_n_lyric_tokens=_A , **_A , ) _UpperCamelCase = version _UpperCamelCase = max_n_lyric_tokens _UpperCamelCase = n_genres with open(_A , encoding='''utf-8''' ) as vocab_handle: _UpperCamelCase = json.load(_A ) with open(_A , encoding='''utf-8''' ) as vocab_handle: _UpperCamelCase = json.load(_A ) with open(_A , encoding='''utf-8''' ) as vocab_handle: _UpperCamelCase = json.load(_A ) _UpperCamelCase = R'''[^A-Za-z0-9.,:;!?\-\'\"()\[\] \t\n]+''' # In v2, we had a n_vocab=80 and in v3 we missed + and so n_vocab=79 of characters. if len(self.lyrics_encoder ) == 79: _UpperCamelCase = oov.replace(R'''\-\'''' , R'''\-+\'''' ) _UpperCamelCase = regex.compile(_A ) _UpperCamelCase = {v: k for k, v in self.artists_encoder.items()} _UpperCamelCase = {v: k for k, v in self.genres_encoder.items()} _UpperCamelCase = {v: k for k, v in self.lyrics_encoder.items()} @property def UpperCamelCase_ ( self : List[str] ): return len(self.artists_encoder ) + len(self.genres_encoder ) + len(self.lyrics_encoder ) def UpperCamelCase_ ( self : Tuple ): return dict(self.artists_encoder , self.genres_encoder , self.lyrics_encoder ) def UpperCamelCase_ ( self : Union[str, Any] , _A : Tuple , _A : Dict , _A : List[Any] ): _UpperCamelCase = [self.artists_encoder.get(_A , 0 ) for artist in list_artists] for genres in range(len(_A ) ): _UpperCamelCase = [self.genres_encoder.get(_A , 0 ) for genre in list_genres[genres]] _UpperCamelCase = list_genres[genres] + [-1] * (self.n_genres - len(list_genres[genres] )) _UpperCamelCase = [[self.lyrics_encoder.get(_A , 0 ) for character in list_lyrics[0]], [], []] return artists_id, list_genres, lyric_ids def UpperCamelCase_ ( self : Tuple , _A : List[Any] ): return list(_A ) def UpperCamelCase_ ( self : Optional[int] , _A : Dict , _A : Optional[Any] , _A : List[str] , **_A : Dict ): _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = self.prepare_for_tokenization(_A , _A , _A ) _UpperCamelCase = self._tokenize(_A ) return artist, genre, lyrics def UpperCamelCase_ ( self : int , _A : str , _A : str , _A : str , _A : bool = False ): for idx in range(len(self.version ) ): if self.version[idx] == "v3": _UpperCamelCase = artists[idx].lower() _UpperCamelCase = [genres[idx].lower()] else: _UpperCamelCase = self._normalize(artists[idx] ) + '''.v2''' _UpperCamelCase = [ self._normalize(_A ) + '''.v2''' for genre in genres[idx].split('''_''' ) ] # split is for the full dictionary with combined genres if self.version[0] == "v2": _UpperCamelCase = regex.compile(R'''[^A-Za-z0-9.,:;!?\-\'\"()\[\] \t\n]+''' ) _UpperCamelCase = '''ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789.,:;!?-+\'\"()[] \t\n''' _UpperCamelCase = {vocab[index]: index + 1 for index in range(len(_A ) )} _UpperCamelCase = 0 _UpperCamelCase = len(_A ) + 1 _UpperCamelCase = self.vocab _UpperCamelCase = {v: k for k, v in self.vocab.items()} _UpperCamelCase = '''''' else: _UpperCamelCase = regex.compile(R'''[^A-Za-z0-9.,:;!?\-+\'\"()\[\] \t\n]+''' ) _UpperCamelCase = self._run_strip_accents(_A ) _UpperCamelCase = lyrics.replace('''\\''' , '''\n''' ) _UpperCamelCase = self.out_of_vocab.sub('''''' , _A ), [], [] return artists, genres, lyrics def UpperCamelCase_ ( self : Optional[int] , _A : Optional[Any] ): _UpperCamelCase = unicodedata.normalize('''NFD''' , _A ) _UpperCamelCase = [] for char in text: _UpperCamelCase = unicodedata.category(_A ) if cat == "Mn": continue output.append(_A ) return "".join(_A ) def UpperCamelCase_ ( self : Tuple , _A : str ): _UpperCamelCase = ( [chr(_A ) for i in range(ord('''a''' ) , ord('''z''' ) + 1 )] + [chr(_A ) for i in range(ord('''A''' ) , ord('''Z''' ) + 1 )] + [chr(_A ) for i in range(ord('''0''' ) , ord('''9''' ) + 1 )] + ['''.'''] ) _UpperCamelCase = frozenset(_A ) _UpperCamelCase = re.compile(R'''_+''' ) _UpperCamelCase = ''''''.join([c if c in accepted else '''_''' for c in text.lower()] ) _UpperCamelCase = pattern.sub('''_''' , _A ).strip('''_''' ) return text def UpperCamelCase_ ( self : Union[str, Any] , _A : List[str] ): return " ".join(_A ) def UpperCamelCase_ ( self : List[str] , _A : List[str] , _A : Optional[Union[str, TensorType]] = None , _A : bool = False ): # Convert to TensorType if not isinstance(_A , _A ): _UpperCamelCase = TensorType(_A ) # Get a function reference for the correct framework if tensor_type == TensorType.TENSORFLOW: if not is_tf_available(): raise ImportError( '''Unable to convert output to TensorFlow tensors format, TensorFlow is not installed.''' ) import tensorflow as tf _UpperCamelCase = tf.constant _UpperCamelCase = tf.is_tensor elif tensor_type == TensorType.PYTORCH: if not is_torch_available(): raise ImportError('''Unable to convert output to PyTorch tensors format, PyTorch is not installed.''' ) import torch _UpperCamelCase = torch.tensor _UpperCamelCase = torch.is_tensor elif tensor_type == TensorType.JAX: if not is_flax_available(): raise ImportError('''Unable to convert output to JAX tensors format, JAX is not installed.''' ) import jax.numpy as jnp # noqa: F811 _UpperCamelCase = jnp.array _UpperCamelCase = _is_jax else: _UpperCamelCase = np.asarray _UpperCamelCase = _is_numpy # Do the tensor conversion in batch try: if prepend_batch_axis: _UpperCamelCase = [inputs] if not is_tensor(_A ): _UpperCamelCase = as_tensor(_A ) except: # noqa E722 raise ValueError( '''Unable to create tensor, you should probably activate truncation and/or padding ''' '''with \'padding=True\' \'truncation=True\' to have batched tensors with the same length.''' ) return inputs def __call__( self : Optional[Any] , _A : Any , _A : int , _A : List[str]="" , _A : Tuple="pt" ): _UpperCamelCase = [0, 0, 0] _UpperCamelCase = [artist] * len(self.version ) _UpperCamelCase = [genres] * len(self.version ) _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = self.tokenize(_A , _A , _A ) _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = self._convert_token_to_id(_A , _A , _A ) _UpperCamelCase = [-INFINITY] * len(full_tokens[-1] ) _UpperCamelCase = [ self.convert_to_tensors( [input_ids + [artists_id[i]] + genres_ids[i] + full_tokens[i]] , tensor_type=_A ) for i in range(len(self.version ) ) ] return BatchEncoding({'''input_ids''': input_ids, '''attention_masks''': attention_masks} ) def UpperCamelCase_ ( self : Optional[Any] , _A : str , _A : Optional[str] = None ): if not os.path.isdir(_A ): logger.error(F"""Vocabulary path ({save_directory}) should be a directory""" ) return _UpperCamelCase = os.path.join( _A , (filename_prefix + '''-''' if filename_prefix else '''''') + VOCAB_FILES_NAMES['''artists_file'''] ) with open(_A , '''w''' , encoding='''utf-8''' ) as f: f.write(json.dumps(self.artists_encoder , ensure_ascii=_A ) ) _UpperCamelCase = os.path.join( _A , (filename_prefix + '''-''' if filename_prefix else '''''') + VOCAB_FILES_NAMES['''genres_file'''] ) with open(_A , '''w''' , encoding='''utf-8''' ) as f: f.write(json.dumps(self.genres_encoder , ensure_ascii=_A ) ) _UpperCamelCase = os.path.join( _A , (filename_prefix + '''-''' if filename_prefix else '''''') + VOCAB_FILES_NAMES['''lyrics_file'''] ) with open(_A , '''w''' , encoding='''utf-8''' ) as f: f.write(json.dumps(self.lyrics_encoder , ensure_ascii=_A ) ) return (artists_file, genres_file, lyrics_file) def UpperCamelCase_ ( self : Dict , _A : Any , _A : int , _A : Any ): _UpperCamelCase = self.artists_decoder.get(_A ) _UpperCamelCase = [self.genres_decoder.get(_A ) for genre in genres_index] _UpperCamelCase = [self.lyrics_decoder.get(_A ) for character in lyric_index] return artist, genres, lyrics
71
def _snake_case ( __snake_case , __snake_case , __snake_case ): if n == 0: return 1 elif n % 2 == 1: return (binary_exponentiation(__snake_case , n - 1 , __snake_case ) * a) % mod else: _UpperCamelCase = binary_exponentiation(__snake_case , n / 2 , __snake_case ) return (b * b) % mod # a prime number _lowerCAmelCase = 701 _lowerCAmelCase = 1_000_000_000 _lowerCAmelCase = 10 # using binary exponentiation function, O(log(p)): print((a / b) % p == (a * binary_exponentiation(b, p - 2, p)) % p) print((a / b) % p == (a * b ** (p - 2)) % p)
71
1
from typing import TYPE_CHECKING from ...utils import ( OptionalDependencyNotAvailable, _LazyModule, is_flax_available, is_sentencepiece_available, is_tf_available, is_tokenizers_available, is_torch_available, ) if is_sentencepiece_available(): from ..ta.tokenization_ta import TaTokenizer else: from ...utils.dummy_sentencepiece_objects import TaTokenizer _lowerCAmelCase = TaTokenizer if is_tokenizers_available(): from ..ta.tokenization_ta_fast import TaTokenizerFast else: from ...utils.dummy_tokenizers_objects import TaTokenizerFast _lowerCAmelCase = TaTokenizerFast _lowerCAmelCase = {"configuration_mt5": ["MT5Config", "MT5OnnxConfig"]} try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "MT5EncoderModel", "MT5ForConditionalGeneration", "MT5ForQuestionAnswering", "MT5Model", "MT5PreTrainedModel", "MT5Stack", ] try: if not is_tf_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = ["TFMT5EncoderModel", "TFMT5ForConditionalGeneration", "TFMT5Model"] try: if not is_flax_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = ["FlaxMT5EncoderModel", "FlaxMT5ForConditionalGeneration", "FlaxMT5Model"] if TYPE_CHECKING: from .configuration_mta import MTaConfig, MTaOnnxConfig try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_mta import ( MTaEncoderModel, MTaForConditionalGeneration, MTaForQuestionAnswering, MTaModel, MTaPreTrainedModel, MTaStack, ) try: if not is_tf_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_tf_mta import TFMTaEncoderModel, TFMTaForConditionalGeneration, TFMTaModel try: if not is_flax_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_flax_mta import FlaxMTaEncoderModel, FlaxMTaForConditionalGeneration, FlaxMTaModel else: import sys _lowerCAmelCase = _LazyModule( __name__, globals()["__file__"], _import_structure, extra_objects={"MT5Tokenizer": MTaTokenizer, "MT5TokenizerFast": MTaTokenizerFast}, module_spec=__spec__, )
71
from math import cos, sin, sqrt, tau from audio_filters.iir_filter import IIRFilter def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 - _cos) / 2 _UpperCamelCase = 1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 + _cos) / 2 _UpperCamelCase = -1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = _sin / 2 _UpperCamelCase = 0 _UpperCamelCase = -ba _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 1 - alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([ba, ba, ba] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = 1 + alpha * big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha * big_a _UpperCamelCase = 1 + alpha / big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha / big_a _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (pmc + aaa) _UpperCamelCase = 2 * big_a * mpc _UpperCamelCase = big_a * (pmc - aaa) _UpperCamelCase = ppmc + aaa _UpperCamelCase = -2 * pmpc _UpperCamelCase = ppmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (ppmc + aaa) _UpperCamelCase = -2 * big_a * pmpc _UpperCamelCase = big_a * (ppmc - aaa) _UpperCamelCase = pmc + aaa _UpperCamelCase = 2 * mpc _UpperCamelCase = pmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt
71
1
import gc import unittest import numpy as np import torch from transformers import CLIPTextConfig, CLIPTextModel, XLMRobertaTokenizer from diffusers import AltDiffusionPipeline, AutoencoderKL, DDIMScheduler, PNDMScheduler, UNetaDConditionModel from diffusers.pipelines.alt_diffusion.modeling_roberta_series import ( RobertaSeriesConfig, RobertaSeriesModelWithTransformation, ) from diffusers.utils import slow, torch_device from diffusers.utils.testing_utils import enable_full_determinism, require_torch_gpu from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS from ..test_pipelines_common import PipelineKarrasSchedulerTesterMixin, PipelineLatentTesterMixin, PipelineTesterMixin enable_full_determinism() class lowerCAmelCase_ ( __lowercase, __lowercase, __lowercase, unittest.TestCase ): UpperCAmelCase = AltDiffusionPipeline UpperCAmelCase = TEXT_TO_IMAGE_PARAMS UpperCAmelCase = TEXT_TO_IMAGE_BATCH_PARAMS UpperCAmelCase = TEXT_TO_IMAGE_IMAGE_PARAMS UpperCAmelCase = TEXT_TO_IMAGE_IMAGE_PARAMS def UpperCamelCase_ ( self : Union[str, Any] ): torch.manual_seed(0 ) _UpperCamelCase = UNetaDConditionModel( block_out_channels=(32, 64) , layers_per_block=2 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=('''DownBlock2D''', '''CrossAttnDownBlock2D''') , up_block_types=('''CrossAttnUpBlock2D''', '''UpBlock2D''') , cross_attention_dim=32 , ) _UpperCamelCase = DDIMScheduler( beta_start=0.0_0085 , beta_end=0.012 , beta_schedule='''scaled_linear''' , clip_sample=_A , set_alpha_to_one=_A , ) torch.manual_seed(0 ) _UpperCamelCase = AutoencoderKL( block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=['''DownEncoderBlock2D''', '''DownEncoderBlock2D'''] , up_block_types=['''UpDecoderBlock2D''', '''UpDecoderBlock2D'''] , latent_channels=4 , ) # TODO: address the non-deterministic text encoder (fails for save-load tests) # torch.manual_seed(0) # text_encoder_config = RobertaSeriesConfig( # hidden_size=32, # project_dim=32, # intermediate_size=37, # layer_norm_eps=1e-05, # num_attention_heads=4, # num_hidden_layers=5, # vocab_size=5002, # ) # text_encoder = RobertaSeriesModelWithTransformation(text_encoder_config) torch.manual_seed(0 ) _UpperCamelCase = CLIPTextConfig( bos_token_id=0 , eos_token_id=2 , hidden_size=32 , projection_dim=32 , intermediate_size=37 , layer_norm_eps=1e-05 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=5002 , ) _UpperCamelCase = CLIPTextModel(_A ) _UpperCamelCase = XLMRobertaTokenizer.from_pretrained('''hf-internal-testing/tiny-xlm-roberta''' ) _UpperCamelCase = 77 _UpperCamelCase = { '''unet''': unet, '''scheduler''': scheduler, '''vae''': vae, '''text_encoder''': text_encoder, '''tokenizer''': tokenizer, '''safety_checker''': None, '''feature_extractor''': None, } return components def UpperCamelCase_ ( self : List[Any] , _A : Tuple , _A : Tuple=0 ): if str(_A ).startswith('''mps''' ): _UpperCamelCase = torch.manual_seed(_A ) else: _UpperCamelCase = torch.Generator(device=_A ).manual_seed(_A ) _UpperCamelCase = { '''prompt''': '''A painting of a squirrel eating a burger''', '''generator''': generator, '''num_inference_steps''': 2, '''guidance_scale''': 6.0, '''output_type''': '''numpy''', } return inputs def UpperCamelCase_ ( self : Optional[int] ): super().test_attention_slicing_forward_pass(expected_max_diff=3e-3 ) def UpperCamelCase_ ( self : Union[str, Any] ): super().test_inference_batch_single_identical(expected_max_diff=3e-3 ) def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = '''cpu''' # ensure determinism for the device-dependent torch.Generator _UpperCamelCase = self.get_dummy_components() torch.manual_seed(0 ) _UpperCamelCase = RobertaSeriesConfig( hidden_size=32 , project_dim=32 , intermediate_size=37 , layer_norm_eps=1e-05 , num_attention_heads=4 , num_hidden_layers=5 , vocab_size=5002 , ) # TODO: remove after fixing the non-deterministic text encoder _UpperCamelCase = RobertaSeriesModelWithTransformation(_A ) _UpperCamelCase = text_encoder _UpperCamelCase = AltDiffusionPipeline(**_A ) _UpperCamelCase = alt_pipe.to(_A ) alt_pipe.set_progress_bar_config(disable=_A ) _UpperCamelCase = self.get_dummy_inputs(_A ) _UpperCamelCase = '''A photo of an astronaut''' _UpperCamelCase = alt_pipe(**_A ) _UpperCamelCase = output.images _UpperCamelCase = image[0, -3:, -3:, -1] assert image.shape == (1, 64, 64, 3) _UpperCamelCase = np.array( [0.574_8162, 0.6044_7145, 0.4882_1217, 0.5010_0636, 0.543_1185, 0.4576_3683, 0.4965_7696, 0.4813_2733, 0.4757_3093] ) assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-2 def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = '''cpu''' # ensure determinism for the device-dependent torch.Generator _UpperCamelCase = self.get_dummy_components() _UpperCamelCase = PNDMScheduler(skip_prk_steps=_A ) torch.manual_seed(0 ) _UpperCamelCase = RobertaSeriesConfig( hidden_size=32 , project_dim=32 , intermediate_size=37 , layer_norm_eps=1e-05 , num_attention_heads=4 , num_hidden_layers=5 , vocab_size=5002 , ) # TODO: remove after fixing the non-deterministic text encoder _UpperCamelCase = RobertaSeriesModelWithTransformation(_A ) _UpperCamelCase = text_encoder _UpperCamelCase = AltDiffusionPipeline(**_A ) _UpperCamelCase = alt_pipe.to(_A ) alt_pipe.set_progress_bar_config(disable=_A ) _UpperCamelCase = self.get_dummy_inputs(_A ) _UpperCamelCase = alt_pipe(**_A ) _UpperCamelCase = output.images _UpperCamelCase = image[0, -3:, -3:, -1] assert image.shape == (1, 64, 64, 3) _UpperCamelCase = np.array( [0.5160_5093, 0.570_7241, 0.4736_5507, 0.5057_8886, 0.563_3877, 0.464_2503, 0.518_2081, 0.4876_3484, 0.4908_4237] ) assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-2 @slow @require_torch_gpu class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : int ): # clean up the VRAM after each test super().tearDown() gc.collect() torch.cuda.empty_cache() def UpperCamelCase_ ( self : Union[str, Any] ): # make sure here that pndm scheduler skips prk _UpperCamelCase = AltDiffusionPipeline.from_pretrained('''BAAI/AltDiffusion''' , safety_checker=_A ) _UpperCamelCase = alt_pipe.to(_A ) alt_pipe.set_progress_bar_config(disable=_A ) _UpperCamelCase = '''A painting of a squirrel eating a burger''' _UpperCamelCase = torch.manual_seed(0 ) _UpperCamelCase = alt_pipe([prompt] , generator=_A , guidance_scale=6.0 , num_inference_steps=20 , output_type='''np''' ) _UpperCamelCase = output.images _UpperCamelCase = image[0, -3:, -3:, -1] assert image.shape == (1, 512, 512, 3) _UpperCamelCase = np.array([0.1010, 0.0800, 0.0794, 0.0885, 0.0843, 0.0762, 0.0769, 0.0729, 0.0586] ) assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-2 def UpperCamelCase_ ( self : int ): _UpperCamelCase = DDIMScheduler.from_pretrained('''BAAI/AltDiffusion''' , subfolder='''scheduler''' ) _UpperCamelCase = AltDiffusionPipeline.from_pretrained('''BAAI/AltDiffusion''' , scheduler=_A , safety_checker=_A ) _UpperCamelCase = alt_pipe.to(_A ) alt_pipe.set_progress_bar_config(disable=_A ) _UpperCamelCase = '''A painting of a squirrel eating a burger''' _UpperCamelCase = torch.manual_seed(0 ) _UpperCamelCase = alt_pipe([prompt] , generator=_A , num_inference_steps=2 , output_type='''numpy''' ) _UpperCamelCase = output.images _UpperCamelCase = image[0, -3:, -3:, -1] assert image.shape == (1, 512, 512, 3) _UpperCamelCase = np.array([0.4019, 0.4052, 0.3810, 0.4119, 0.3916, 0.3982, 0.4651, 0.4195, 0.5323] ) assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-2
71
from ...configuration_utils import PretrainedConfig from ...utils import logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "EleutherAI/gpt-neox-20b": "https://huggingface.co/EleutherAI/gpt-neox-20b/resolve/main/config.json", # See all GPTNeoX models at https://huggingface.co/models?filter=gpt_neox } class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "gpt_neox" def __init__( self : Union[str, Any] , _A : Union[str, Any]=5_0432 , _A : List[Any]=6144 , _A : int=44 , _A : int=64 , _A : Optional[Any]=2_4576 , _A : Any="gelu" , _A : Tuple=0.25 , _A : Union[str, Any]=1_0000 , _A : Tuple=0.0 , _A : Any=0.0 , _A : int=0.1 , _A : List[str]=2048 , _A : Dict=0.02 , _A : Optional[Any]=1e-5 , _A : Tuple=True , _A : List[Any]=0 , _A : Optional[int]=2 , _A : Optional[int]=False , _A : List[Any]=True , _A : Any=None , **_A : Any , ): super().__init__(bos_token_id=_A , eos_token_id=_A , **_A ) _UpperCamelCase = vocab_size _UpperCamelCase = max_position_embeddings _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = rotary_pct _UpperCamelCase = rotary_emb_base _UpperCamelCase = attention_dropout _UpperCamelCase = hidden_dropout _UpperCamelCase = classifier_dropout _UpperCamelCase = initializer_range _UpperCamelCase = layer_norm_eps _UpperCamelCase = use_cache _UpperCamelCase = tie_word_embeddings _UpperCamelCase = use_parallel_residual _UpperCamelCase = rope_scaling self._rope_scaling_validation() if self.hidden_size % self.num_attention_heads != 0: raise ValueError( '''The hidden size is not divisble by the number of attention heads! Make sure to update them!''' ) def UpperCamelCase_ ( self : str ): if self.rope_scaling is None: return if not isinstance(self.rope_scaling , _A ) or len(self.rope_scaling ) != 2: raise ValueError( '''`rope_scaling` must be a dictionary with with two fields, `name` and `factor`, ''' F"""got {self.rope_scaling}""" ) _UpperCamelCase = self.rope_scaling.get('''type''' , _A ) _UpperCamelCase = self.rope_scaling.get('''factor''' , _A ) if rope_scaling_type is None or rope_scaling_type not in ["linear", "dynamic"]: raise ValueError( F"""`rope_scaling`'s name field must be one of ['linear', 'dynamic'], got {rope_scaling_type}""" ) if rope_scaling_factor is None or not isinstance(_A , _A ) or rope_scaling_factor <= 1.0: raise ValueError(F"""`rope_scaling`'s factor field must be an float > 1, got {rope_scaling_factor}""" )
71
1
import re def _snake_case ( __snake_case ): return [char.split() for char in re.split(R'''[^ a-z A-Z 0-9 \s]''' , str_ )] def _snake_case ( __snake_case ): _UpperCamelCase = split_input(str_ ) return "".join( [''''''.join([char.capitalize() for char in sub_str] ) for sub_str in string_split] ) def _snake_case ( __snake_case , __snake_case , __snake_case ): try: _UpperCamelCase = split_input(__snake_case ) if upper: _UpperCamelCase = ''''''.join( [ separator.join([char.upper() for char in sub_str] ) for sub_str in string_split ] ) else: _UpperCamelCase = ''''''.join( [ separator.join([char.lower() for char in sub_str] ) for sub_str in string_split ] ) return res_str except IndexError: return "not valid string" def _snake_case ( __snake_case ): return to_simple_case(__snake_case ) def _snake_case ( __snake_case ): try: _UpperCamelCase = to_simple_case(__snake_case ) return res_str[0].lower() + res_str[1:] except IndexError: return "not valid string" def _snake_case ( __snake_case , __snake_case ): return to_complex_case(__snake_case , __snake_case , '''_''' ) def _snake_case ( __snake_case , __snake_case ): return to_complex_case(__snake_case , __snake_case , '''-''' ) if __name__ == "__main__": __import__("doctest").testmod()
71
from ..utils import DummyObject, requires_backends class lowerCAmelCase_ ( metaclass=__lowercase ): UpperCAmelCase = ["keras_nlp"] def __init__( self : Any , *_A : Dict , **_A : List[str] ): requires_backends(self , ['''keras_nlp'''] )
71
1
def _snake_case ( __snake_case ): _UpperCamelCase = 0 _UpperCamelCase = len(__snake_case ) for i in range(n - 1 ): for j in range(i + 1 , __snake_case ): if arr[i] > arr[j]: num_inversions += 1 return num_inversions def _snake_case ( __snake_case ): if len(__snake_case ) <= 1: return arr, 0 _UpperCamelCase = len(__snake_case ) // 2 _UpperCamelCase = arr[0:mid] _UpperCamelCase = arr[mid:] _UpperCamelCase , _UpperCamelCase = count_inversions_recursive(__snake_case ) _UpperCamelCase , _UpperCamelCase = count_inversions_recursive(__snake_case ) _UpperCamelCase , _UpperCamelCase = _count_cross_inversions(__snake_case , __snake_case ) _UpperCamelCase = inversion_p + inversions_q + cross_inversions return c, num_inversions def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = [] _UpperCamelCase = _UpperCamelCase = _UpperCamelCase = 0 while i < len(__snake_case ) and j < len(__snake_case ): if p[i] > q[j]: # if P[1] > Q[j], then P[k] > Q[k] for all i < k <= len(P) # These are all inversions. The claim emerges from the # property that P is sorted. num_inversion += len(__snake_case ) - i r.append(q[j] ) j += 1 else: r.append(p[i] ) i += 1 if i < len(__snake_case ): r.extend(p[i:] ) else: r.extend(q[j:] ) return r, num_inversion def _snake_case ( ): _UpperCamelCase = [10, 2, 1, 5, 5, 2, 11] # this arr has 8 inversions: # (10, 2), (10, 1), (10, 5), (10, 5), (10, 2), (2, 1), (5, 2), (5, 2) _UpperCamelCase = count_inversions_bf(__snake_case ) _UpperCamelCase , _UpperCamelCase = count_inversions_recursive(__snake_case ) assert num_inversions_bf == num_inversions_recursive == 8 print('''number of inversions = ''' , __snake_case ) # testing an array with zero inversion (a sorted arr_1) arr_a.sort() _UpperCamelCase = count_inversions_bf(__snake_case ) _UpperCamelCase , _UpperCamelCase = count_inversions_recursive(__snake_case ) assert num_inversions_bf == num_inversions_recursive == 0 print('''number of inversions = ''' , __snake_case ) # an empty list should also have zero inversions _UpperCamelCase = [] _UpperCamelCase = count_inversions_bf(__snake_case ) _UpperCamelCase , _UpperCamelCase = count_inversions_recursive(__snake_case ) assert num_inversions_bf == num_inversions_recursive == 0 print('''number of inversions = ''' , __snake_case ) if __name__ == "__main__": main()
71
from typing import Optional, Tuple, Union import tensorflow as tf from ...activations_tf import ACTaFN from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward from ...modeling_tf_outputs import ( TFBaseModelOutputWithNoAttention, TFBaseModelOutputWithPoolingAndNoAttention, TFSequenceClassifierOutput, ) from ...modeling_tf_utils import TFPreTrainedModel, TFSequenceClassificationLoss, keras_serializable, unpack_inputs from ...tf_utils import shape_list from ...utils import logging from .configuration_regnet import RegNetConfig _lowerCAmelCase = logging.get_logger(__name__) # General docstring _lowerCAmelCase = "RegNetConfig" # Base docstring _lowerCAmelCase = "facebook/regnet-y-040" _lowerCAmelCase = [1, 1_088, 7, 7] # Image classification docstring _lowerCAmelCase = "facebook/regnet-y-040" _lowerCAmelCase = "tabby, tabby cat" _lowerCAmelCase = [ "facebook/regnet-y-040", # See all regnet models at https://huggingface.co/models?filter=regnet ] class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : str , _A : int , _A : int = 3 , _A : int = 1 , _A : int = 1 , _A : Optional[str] = "relu" , **_A : Any , ): super().__init__(**_A ) # The padding and conv has been verified in # https://colab.research.google.com/gist/sayakpaul/854bc10eeaf21c9ee2119e0b9f3841a7/scratchpad.ipynb _UpperCamelCase = tf.keras.layers.ZeroPaddingaD(padding=kernel_size // 2 ) _UpperCamelCase = tf.keras.layers.ConvaD( filters=_A , kernel_size=_A , strides=_A , padding='''VALID''' , groups=_A , use_bias=_A , name='''convolution''' , ) _UpperCamelCase = tf.keras.layers.BatchNormalization(epsilon=1e-5 , momentum=0.9 , name='''normalization''' ) _UpperCamelCase = ACTaFN[activation] if activation is not None else tf.identity def UpperCamelCase_ ( self : Any , _A : Any ): _UpperCamelCase = self.convolution(self.padding(_A ) ) _UpperCamelCase = self.normalization(_A ) _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Optional[Any] , _A : RegNetConfig , **_A : Any ): super().__init__(**_A ) _UpperCamelCase = config.num_channels _UpperCamelCase = TFRegNetConvLayer( out_channels=config.embedding_size , kernel_size=3 , stride=2 , activation=config.hidden_act , name='''embedder''' , ) def UpperCamelCase_ ( self : List[str] , _A : Optional[int] ): _UpperCamelCase = shape_list(_A )[1] if tf.executing_eagerly() and num_channels != self.num_channels: raise ValueError( '''Make sure that the channel dimension of the pixel values match with the one set in the configuration.''' ) # When running on CPU, `tf.keras.layers.Conv2D` doesn't support `NCHW` format. # So change the input format from `NCHW` to `NHWC`. # shape = (batch_size, in_height, in_width, in_channels=num_channels) _UpperCamelCase = tf.transpose(_A , perm=(0, 2, 3, 1) ) _UpperCamelCase = self.embedder(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : str , _A : int , _A : int = 2 , **_A : Optional[Any] ): super().__init__(**_A ) _UpperCamelCase = tf.keras.layers.ConvaD( filters=_A , kernel_size=1 , strides=_A , use_bias=_A , name='''convolution''' ) _UpperCamelCase = tf.keras.layers.BatchNormalization(epsilon=1e-5 , momentum=0.9 , name='''normalization''' ) def UpperCamelCase_ ( self : str , _A : tf.Tensor , _A : bool = False ): return self.normalization(self.convolution(_A ) , training=_A ) class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Dict , _A : int , _A : int , **_A : Dict ): super().__init__(**_A ) _UpperCamelCase = tf.keras.layers.GlobalAveragePoolingaD(keepdims=_A , name='''pooler''' ) _UpperCamelCase = [ tf.keras.layers.ConvaD(filters=_A , kernel_size=1 , activation='''relu''' , name='''attention.0''' ), tf.keras.layers.ConvaD(filters=_A , kernel_size=1 , activation='''sigmoid''' , name='''attention.2''' ), ] def UpperCamelCase_ ( self : List[str] , _A : List[Any] ): # [batch_size, h, w, num_channels] -> [batch_size, 1, 1, num_channels] _UpperCamelCase = self.pooler(_A ) for layer_module in self.attention: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = hidden_state * pooled return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , _A : int , _A : int , _A : int = 1 , **_A : str ): super().__init__(**_A ) _UpperCamelCase = in_channels != out_channels or stride != 1 _UpperCamelCase = max(1 , out_channels // config.groups_width ) _UpperCamelCase = ( TFRegNetShortCut(_A , stride=_A , name='''shortcut''' ) if should_apply_shortcut else tf.keras.layers.Activation('''linear''' , name='''shortcut''' ) ) # `self.layers` instead of `self.layer` because that is a reserved argument. _UpperCamelCase = [ TFRegNetConvLayer(_A , kernel_size=1 , activation=config.hidden_act , name='''layer.0''' ), TFRegNetConvLayer( _A , stride=_A , groups=_A , activation=config.hidden_act , name='''layer.1''' ), TFRegNetConvLayer(_A , kernel_size=1 , activation=_A , name='''layer.2''' ), ] _UpperCamelCase = ACTaFN[config.hidden_act] def UpperCamelCase_ ( self : Dict , _A : Tuple ): _UpperCamelCase = hidden_state for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = self.shortcut(_A ) hidden_state += residual _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , _A : int , _A : int , _A : int = 1 , **_A : int ): super().__init__(**_A ) _UpperCamelCase = in_channels != out_channels or stride != 1 _UpperCamelCase = max(1 , out_channels // config.groups_width ) _UpperCamelCase = ( TFRegNetShortCut(_A , stride=_A , name='''shortcut''' ) if should_apply_shortcut else tf.keras.layers.Activation('''linear''' , name='''shortcut''' ) ) _UpperCamelCase = [ TFRegNetConvLayer(_A , kernel_size=1 , activation=config.hidden_act , name='''layer.0''' ), TFRegNetConvLayer( _A , stride=_A , groups=_A , activation=config.hidden_act , name='''layer.1''' ), TFRegNetSELayer(_A , reduced_channels=int(round(in_channels / 4 ) ) , name='''layer.2''' ), TFRegNetConvLayer(_A , kernel_size=1 , activation=_A , name='''layer.3''' ), ] _UpperCamelCase = ACTaFN[config.hidden_act] def UpperCamelCase_ ( self : Tuple , _A : List[Any] ): _UpperCamelCase = hidden_state for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = self.shortcut(_A ) hidden_state += residual _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Tuple , _A : RegNetConfig , _A : int , _A : int , _A : int = 2 , _A : int = 2 , **_A : Union[str, Any] ): super().__init__(**_A ) _UpperCamelCase = TFRegNetXLayer if config.layer_type == '''x''' else TFRegNetYLayer _UpperCamelCase = [ # downsampling is done in the first layer with stride of 2 layer(_A , _A , _A , stride=_A , name='''layers.0''' ), *[layer(_A , _A , _A , name=F"""layers.{i+1}""" ) for i in range(depth - 1 )], ] def UpperCamelCase_ ( self : Union[str, Any] , _A : Optional[int] ): for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , **_A : List[str] ): super().__init__(**_A ) _UpperCamelCase = [] # based on `downsample_in_first_stage`, the first layer of the first stage may or may not downsample the input self.stages.append( TFRegNetStage( _A , config.embedding_size , config.hidden_sizes[0] , stride=2 if config.downsample_in_first_stage else 1 , depth=config.depths[0] , name='''stages.0''' , ) ) _UpperCamelCase = zip(config.hidden_sizes , config.hidden_sizes[1:] ) for i, ((in_channels, out_channels), depth) in enumerate(zip(_A , config.depths[1:] ) ): self.stages.append(TFRegNetStage(_A , _A , _A , depth=_A , name=F"""stages.{i+1}""" ) ) def UpperCamelCase_ ( self : Optional[int] , _A : tf.Tensor , _A : bool = False , _A : bool = True ): _UpperCamelCase = () if output_hidden_states else None for stage_module in self.stages: if output_hidden_states: _UpperCamelCase = hidden_states + (hidden_state,) _UpperCamelCase = stage_module(_A ) if output_hidden_states: _UpperCamelCase = hidden_states + (hidden_state,) if not return_dict: return tuple(v for v in [hidden_state, hidden_states] if v is not None ) return TFBaseModelOutputWithNoAttention(last_hidden_state=_A , hidden_states=_A ) @keras_serializable class lowerCAmelCase_ ( tf.keras.layers.Layer ): UpperCAmelCase = RegNetConfig def __init__( self : int , _A : Tuple , **_A : int ): super().__init__(**_A ) _UpperCamelCase = config _UpperCamelCase = TFRegNetEmbeddings(_A , name='''embedder''' ) _UpperCamelCase = TFRegNetEncoder(_A , name='''encoder''' ) _UpperCamelCase = tf.keras.layers.GlobalAveragePoolingaD(keepdims=_A , name='''pooler''' ) @unpack_inputs def UpperCamelCase_ ( self : Optional[int] , _A : tf.Tensor , _A : Optional[bool] = None , _A : Optional[bool] = None , _A : bool = False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.embedder(_A , training=_A ) _UpperCamelCase = self.encoder( _A , output_hidden_states=_A , return_dict=_A , training=_A ) _UpperCamelCase = encoder_outputs[0] _UpperCamelCase = self.pooler(_A ) # Change to NCHW output format have uniformity in the modules _UpperCamelCase = tf.transpose(_A , perm=(0, 3, 1, 2) ) _UpperCamelCase = tf.transpose(_A , perm=(0, 3, 1, 2) ) # Change the other hidden state outputs to NCHW as well if output_hidden_states: _UpperCamelCase = tuple([tf.transpose(_A , perm=(0, 3, 1, 2) ) for h in encoder_outputs[1]] ) if not return_dict: return (last_hidden_state, pooled_output) + encoder_outputs[1:] return TFBaseModelOutputWithPoolingAndNoAttention( last_hidden_state=_A , pooler_output=_A , hidden_states=hidden_states if output_hidden_states else encoder_outputs.hidden_states , ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = RegNetConfig UpperCAmelCase = "regnet" UpperCAmelCase = "pixel_values" @property def UpperCamelCase_ ( self : Tuple ): return {"pixel_values": tf.TensorSpec(shape=(None, self.config.num_channels, 224, 224) , dtype=tf.floataa )} _lowerCAmelCase = r"\n Parameters:\n This model is a Tensorflow\n [tf.keras.layers.Layer](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Layer) sub-class. Use it as a\n regular Tensorflow Module and refer to the Tensorflow documentation for all matter related to general usage and\n behavior.\n config ([`RegNetConfig`]): Model configuration class with all the parameters of the model.\n Initializing with a config file does not load the weights associated with the model, only the\n configuration. Check out the [`~TFPreTrainedModel.from_pretrained`] method to load the model weights.\n" _lowerCAmelCase = r"\n Args:\n pixel_values (`tf.Tensor` of shape `(batch_size, num_channels, height, width)`):\n Pixel values. Pixel values can be obtained using [`AutoImageProcessor`]. See\n [`ConveNextImageProcessor.__call__`] for details.\n output_hidden_states (`bool`, *optional*):\n Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for\n more detail.\n return_dict (`bool`, *optional*):\n Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.\n" @add_start_docstrings( "The bare RegNet model outputting raw features without any specific head on top.", __lowercase, ) class lowerCAmelCase_ ( __lowercase ): def __init__( self : List[Any] , _A : RegNetConfig , *_A : Optional[int] , **_A : Tuple ): super().__init__(_A , *_A , **_A ) _UpperCamelCase = TFRegNetMainLayer(_A , name='''regnet''' ) @unpack_inputs @add_start_docstrings_to_model_forward(_A ) @add_code_sample_docstrings( checkpoint=_CHECKPOINT_FOR_DOC , output_type=_A , config_class=_CONFIG_FOR_DOC , modality='''vision''' , expected_output=_EXPECTED_OUTPUT_SHAPE , ) def UpperCamelCase_ ( self : Any , _A : tf.Tensor , _A : Optional[bool] = None , _A : Optional[bool] = None , _A : Optional[int]=False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.regnet( pixel_values=_A , output_hidden_states=_A , return_dict=_A , training=_A , ) if not return_dict: return (outputs[0],) + outputs[1:] return TFBaseModelOutputWithPoolingAndNoAttention( last_hidden_state=outputs.last_hidden_state , pooler_output=outputs.pooler_output , hidden_states=outputs.hidden_states , ) @add_start_docstrings( "\n RegNet Model with an image classification head on top (a linear layer on top of the pooled features), e.g. for\n ImageNet.\n ", __lowercase, ) class lowerCAmelCase_ ( __lowercase, __lowercase ): def __init__( self : List[Any] , _A : RegNetConfig , *_A : Any , **_A : int ): super().__init__(_A , *_A , **_A ) _UpperCamelCase = config.num_labels _UpperCamelCase = TFRegNetMainLayer(_A , name='''regnet''' ) # classification head _UpperCamelCase = [ tf.keras.layers.Flatten(), tf.keras.layers.Dense(config.num_labels , name='''classifier.1''' ) if config.num_labels > 0 else tf.identity, ] @unpack_inputs @add_start_docstrings_to_model_forward(_A ) @add_code_sample_docstrings( checkpoint=_IMAGE_CLASS_CHECKPOINT , output_type=_A , config_class=_CONFIG_FOR_DOC , expected_output=_IMAGE_CLASS_EXPECTED_OUTPUT , ) def UpperCamelCase_ ( self : str , _A : tf.Tensor = None , _A : tf.Tensor = None , _A : bool = None , _A : bool = None , _A : Any=False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.regnet( _A , output_hidden_states=_A , return_dict=_A , training=_A ) _UpperCamelCase = outputs.pooler_output if return_dict else outputs[1] _UpperCamelCase = self.classifier[0](_A ) _UpperCamelCase = self.classifier[1](_A ) _UpperCamelCase = None if labels is None else self.hf_compute_loss(labels=_A , logits=_A ) if not return_dict: _UpperCamelCase = (logits,) + outputs[2:] return ((loss,) + output) if loss is not None else output return TFSequenceClassifierOutput(loss=_A , logits=_A , hidden_states=outputs.hidden_states )
71
1
import argparse import io import requests import torch from omegaconf import OmegaConf from diffusers import AutoencoderKL from diffusers.pipelines.stable_diffusion.convert_from_ckpt import ( assign_to_checkpoint, conv_attn_to_linear, create_vae_diffusers_config, renew_vae_attention_paths, renew_vae_resnet_paths, ) def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = checkpoint _UpperCamelCase = {} _UpperCamelCase = vae_state_dict['''encoder.conv_in.weight'''] _UpperCamelCase = vae_state_dict['''encoder.conv_in.bias'''] _UpperCamelCase = vae_state_dict['''encoder.conv_out.weight'''] _UpperCamelCase = vae_state_dict['''encoder.conv_out.bias'''] _UpperCamelCase = vae_state_dict['''encoder.norm_out.weight'''] _UpperCamelCase = vae_state_dict['''encoder.norm_out.bias'''] _UpperCamelCase = vae_state_dict['''decoder.conv_in.weight'''] _UpperCamelCase = vae_state_dict['''decoder.conv_in.bias'''] _UpperCamelCase = vae_state_dict['''decoder.conv_out.weight'''] _UpperCamelCase = vae_state_dict['''decoder.conv_out.bias'''] _UpperCamelCase = vae_state_dict['''decoder.norm_out.weight'''] _UpperCamelCase = vae_state_dict['''decoder.norm_out.bias'''] _UpperCamelCase = vae_state_dict['''quant_conv.weight'''] _UpperCamelCase = vae_state_dict['''quant_conv.bias'''] _UpperCamelCase = vae_state_dict['''post_quant_conv.weight'''] _UpperCamelCase = vae_state_dict['''post_quant_conv.bias'''] # Retrieves the keys for the encoder down blocks only _UpperCamelCase = len({'''.'''.join(layer.split('''.''' )[:3] ) for layer in vae_state_dict if '''encoder.down''' in layer} ) _UpperCamelCase = { layer_id: [key for key in vae_state_dict if f"""down.{layer_id}""" in key] for layer_id in range(__snake_case ) } # Retrieves the keys for the decoder up blocks only _UpperCamelCase = len({'''.'''.join(layer.split('''.''' )[:3] ) for layer in vae_state_dict if '''decoder.up''' in layer} ) _UpperCamelCase = { layer_id: [key for key in vae_state_dict if f"""up.{layer_id}""" in key] for layer_id in range(__snake_case ) } for i in range(__snake_case ): _UpperCamelCase = [key for key in down_blocks[i] if f"""down.{i}""" in key and f"""down.{i}.downsample""" not in key] if f"""encoder.down.{i}.downsample.conv.weight""" in vae_state_dict: _UpperCamelCase = vae_state_dict.pop( f"""encoder.down.{i}.downsample.conv.weight""" ) _UpperCamelCase = vae_state_dict.pop( f"""encoder.down.{i}.downsample.conv.bias""" ) _UpperCamelCase = renew_vae_resnet_paths(__snake_case ) _UpperCamelCase = {'''old''': f"""down.{i}.block""", '''new''': f"""down_blocks.{i}.resnets"""} assign_to_checkpoint(__snake_case , __snake_case , __snake_case , additional_replacements=[meta_path] , config=__snake_case ) _UpperCamelCase = [key for key in vae_state_dict if '''encoder.mid.block''' in key] _UpperCamelCase = 2 for i in range(1 , num_mid_res_blocks + 1 ): _UpperCamelCase = [key for key in mid_resnets if f"""encoder.mid.block_{i}""" in key] _UpperCamelCase = renew_vae_resnet_paths(__snake_case ) _UpperCamelCase = {'''old''': f"""mid.block_{i}""", '''new''': f"""mid_block.resnets.{i - 1}"""} assign_to_checkpoint(__snake_case , __snake_case , __snake_case , additional_replacements=[meta_path] , config=__snake_case ) _UpperCamelCase = [key for key in vae_state_dict if '''encoder.mid.attn''' in key] _UpperCamelCase = renew_vae_attention_paths(__snake_case ) _UpperCamelCase = {'''old''': '''mid.attn_1''', '''new''': '''mid_block.attentions.0'''} assign_to_checkpoint(__snake_case , __snake_case , __snake_case , additional_replacements=[meta_path] , config=__snake_case ) conv_attn_to_linear(__snake_case ) for i in range(__snake_case ): _UpperCamelCase = num_up_blocks - 1 - i _UpperCamelCase = [ key for key in up_blocks[block_id] if f"""up.{block_id}""" in key and f"""up.{block_id}.upsample""" not in key ] if f"""decoder.up.{block_id}.upsample.conv.weight""" in vae_state_dict: _UpperCamelCase = vae_state_dict[ f"""decoder.up.{block_id}.upsample.conv.weight""" ] _UpperCamelCase = vae_state_dict[ f"""decoder.up.{block_id}.upsample.conv.bias""" ] _UpperCamelCase = renew_vae_resnet_paths(__snake_case ) _UpperCamelCase = {'''old''': f"""up.{block_id}.block""", '''new''': f"""up_blocks.{i}.resnets"""} assign_to_checkpoint(__snake_case , __snake_case , __snake_case , additional_replacements=[meta_path] , config=__snake_case ) _UpperCamelCase = [key for key in vae_state_dict if '''decoder.mid.block''' in key] _UpperCamelCase = 2 for i in range(1 , num_mid_res_blocks + 1 ): _UpperCamelCase = [key for key in mid_resnets if f"""decoder.mid.block_{i}""" in key] _UpperCamelCase = renew_vae_resnet_paths(__snake_case ) _UpperCamelCase = {'''old''': f"""mid.block_{i}""", '''new''': f"""mid_block.resnets.{i - 1}"""} assign_to_checkpoint(__snake_case , __snake_case , __snake_case , additional_replacements=[meta_path] , config=__snake_case ) _UpperCamelCase = [key for key in vae_state_dict if '''decoder.mid.attn''' in key] _UpperCamelCase = renew_vae_attention_paths(__snake_case ) _UpperCamelCase = {'''old''': '''mid.attn_1''', '''new''': '''mid_block.attentions.0'''} assign_to_checkpoint(__snake_case , __snake_case , __snake_case , additional_replacements=[meta_path] , config=__snake_case ) conv_attn_to_linear(__snake_case ) return new_checkpoint def _snake_case ( __snake_case , __snake_case , ): # Only support V1 _UpperCamelCase = requests.get( ''' https://raw.githubusercontent.com/CompVis/stable-diffusion/main/configs/stable-diffusion/v1-inference.yaml''' ) _UpperCamelCase = io.BytesIO(r.content ) _UpperCamelCase = OmegaConf.load(__snake_case ) _UpperCamelCase = 512 _UpperCamelCase = '''cuda''' if torch.cuda.is_available() else '''cpu''' if checkpoint_path.endswith('''safetensors''' ): from safetensors import safe_open _UpperCamelCase = {} with safe_open(__snake_case , framework='''pt''' , device='''cpu''' ) as f: for key in f.keys(): _UpperCamelCase = f.get_tensor(__snake_case ) else: _UpperCamelCase = torch.load(__snake_case , map_location=__snake_case )['''state_dict'''] # Convert the VAE model. _UpperCamelCase = create_vae_diffusers_config(__snake_case , image_size=__snake_case ) _UpperCamelCase = custom_convert_ldm_vae_checkpoint(__snake_case , __snake_case ) _UpperCamelCase = AutoencoderKL(**__snake_case ) vae.load_state_dict(__snake_case ) vae.save_pretrained(__snake_case ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() parser.add_argument("--vae_pt_path", default=None, type=str, required=True, help="Path to the VAE.pt to convert.") parser.add_argument("--dump_path", default=None, type=str, required=True, help="Path to the VAE.pt to convert.") _lowerCAmelCase = parser.parse_args() vae_pt_to_vae_diffuser(args.vae_pt_path, args.dump_path)
71
from sklearn.metrics import mean_squared_error import datasets _lowerCAmelCase = "\\n@article{scikit-learn,\n title={Scikit-learn: Machine Learning in {P}ython},\n author={Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V.\n and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P.\n and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and\n Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E.},\n journal={Journal of Machine Learning Research},\n volume={12},\n pages={2825--2830},\n year={2011}\n}\n" _lowerCAmelCase = "\\nMean Squared Error(MSE) is the average of the square of difference between the predicted\nand actual values.\n" _lowerCAmelCase = "\nArgs:\n predictions: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Estimated target values.\n references: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Ground truth (correct) target values.\n sample_weight: array-like of shape (n_samples,), default=None\n Sample weights.\n multioutput: {\"raw_values\", \"uniform_average\"} or array-like of shape (n_outputs,), default=\"uniform_average\"\n Defines aggregating of multiple output values. Array-like value defines weights used to average errors.\n\n \"raw_values\" : Returns a full set of errors in case of multioutput input.\n\n \"uniform_average\" : Errors of all outputs are averaged with uniform weight.\n\n squared : bool, default=True\n If True returns MSE value, if False returns RMSE (Root Mean Squared Error) value.\n\nReturns:\n mse : mean squared error.\nExamples:\n\n >>> mse_metric = datasets.load_metric(\"mse\")\n >>> predictions = [2.5, 0.0, 2, 8]\n >>> references = [3, -0.5, 2, 7]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.375}\n >>> rmse_result = mse_metric.compute(predictions=predictions, references=references, squared=False)\n >>> print(rmse_result)\n {'mse': 0.6123724356957945}\n\n If you're using multi-dimensional lists, then set the config as follows :\n\n >>> mse_metric = datasets.load_metric(\"mse\", \"multilist\")\n >>> predictions = [[0.5, 1], [-1, 1], [7, -6]]\n >>> references = [[0, 2], [-1, 2], [8, -5]]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.7083333333333334}\n >>> results = mse_metric.compute(predictions=predictions, references=references, multioutput='raw_values')\n >>> print(results) # doctest: +NORMALIZE_WHITESPACE\n {'mse': array([0.41666667, 1. ])}\n" @datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION ) class lowerCAmelCase_ ( datasets.Metric ): def UpperCamelCase_ ( self : Optional[int] ): return datasets.MetricInfo( description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features(self._get_feature_types() ) , reference_urls=[ '''https://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html''' ] , ) def UpperCamelCase_ ( self : Dict ): if self.config_name == "multilist": return { "predictions": datasets.Sequence(datasets.Value('''float''' ) ), "references": datasets.Sequence(datasets.Value('''float''' ) ), } else: return { "predictions": datasets.Value('''float''' ), "references": datasets.Value('''float''' ), } def UpperCamelCase_ ( self : Any , _A : List[Any] , _A : List[str] , _A : Dict=None , _A : List[str]="uniform_average" , _A : int=True ): _UpperCamelCase = mean_squared_error( _A , _A , sample_weight=_A , multioutput=_A , squared=_A ) return {"mse": mse}
71
1
import argparse import shutil from pathlib import Path from tqdm import tqdm from transformers import AutoTokenizer def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case=1024 ): _UpperCamelCase , _UpperCamelCase = [], [] _UpperCamelCase = list(zip(__snake_case , __snake_case ) ) _UpperCamelCase , _UpperCamelCase = sorted_examples[0] def is_too_big(__snake_case ): return tok(__snake_case , return_tensors='''pt''' ).input_ids.shape[1] > max_tokens for src, tgt in tqdm(sorted_examples[1:] ): _UpperCamelCase = new_src + ''' ''' + src _UpperCamelCase = new_tgt + ''' ''' + tgt if is_too_big(__snake_case ) or is_too_big(__snake_case ): # cant fit, finalize example finished_src.append(__snake_case ) finished_tgt.append(__snake_case ) _UpperCamelCase , _UpperCamelCase = src, tgt else: # can fit, keep adding _UpperCamelCase , _UpperCamelCase = cand_src, cand_tgt # cleanup if new_src: assert new_tgt finished_src.append(__snake_case ) finished_tgt.append(__snake_case ) return finished_src, finished_tgt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = Path(__snake_case ) save_path.mkdir(exist_ok=__snake_case ) for split in ["train"]: _UpperCamelCase , _UpperCamelCase = data_dir / f"""{split}.source""", data_dir / f"""{split}.target""" _UpperCamelCase = [x.rstrip() for x in Path(__snake_case ).open().readlines()] _UpperCamelCase = [x.rstrip() for x in Path(__snake_case ).open().readlines()] _UpperCamelCase , _UpperCamelCase = pack_examples(__snake_case , __snake_case , __snake_case , __snake_case ) print(f"""packed {split} split from {len(__snake_case )} examples -> {len(__snake_case )}.""" ) Path(save_path / f"""{split}.source""" ).open('''w''' ).write('''\n'''.join(__snake_case ) ) Path(save_path / f"""{split}.target""" ).open('''w''' ).write('''\n'''.join(__snake_case ) ) for split in ["val", "test"]: _UpperCamelCase , _UpperCamelCase = data_dir / f"""{split}.source""", data_dir / f"""{split}.target""" shutil.copyfile(__snake_case , save_path / f"""{split}.source""" ) shutil.copyfile(__snake_case , save_path / f"""{split}.target""" ) def _snake_case ( ): _UpperCamelCase = argparse.ArgumentParser() parser.add_argument('''--tok_name''' , type=__snake_case , help='''like facebook/bart-large-cnn,t5-base, etc.''' ) parser.add_argument('''--max_seq_len''' , type=__snake_case , default=128 ) parser.add_argument('''--data_dir''' , type=__snake_case ) parser.add_argument('''--save_path''' , type=__snake_case ) _UpperCamelCase = parser.parse_args() _UpperCamelCase = AutoTokenizer.from_pretrained(args.tok_name ) return pack_data_dir(__snake_case , Path(args.data_dir ) , args.max_seq_len , args.save_path ) if __name__ == "__main__": packer_cli()
71
import os import re import shutil import sys import tempfile import unittest import black _lowerCAmelCase = os.path.abspath(os.path.dirname(os.path.dirname(os.path.dirname(__file__)))) sys.path.append(os.path.join(git_repo_path, "utils")) import check_copies # noqa: E402 # This is the reference code that will be used in the tests. # If DDPMSchedulerOutput is changed in scheduling_ddpm.py, this code needs to be manually updated. _lowerCAmelCase = " \"\"\"\n Output class for the scheduler's step function output.\n\n Args:\n prev_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n Computed sample (x_{t-1}) of previous timestep. `prev_sample` should be used as next model input in the\n denoising loop.\n pred_original_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n The predicted denoised sample (x_{0}) based on the model output from the current timestep.\n `pred_original_sample` can be used to preview progress or for guidance.\n \"\"\"\n\n prev_sample: torch.FloatTensor\n pred_original_sample: Optional[torch.FloatTensor] = None\n" class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = tempfile.mkdtemp() os.makedirs(os.path.join(self.diffusers_dir , '''schedulers/''' ) ) _UpperCamelCase = self.diffusers_dir shutil.copy( os.path.join(_A , '''src/diffusers/schedulers/scheduling_ddpm.py''' ) , os.path.join(self.diffusers_dir , '''schedulers/scheduling_ddpm.py''' ) , ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = '''src/diffusers''' shutil.rmtree(self.diffusers_dir ) def UpperCamelCase_ ( self : Union[str, Any] , _A : Tuple , _A : Optional[Any] , _A : Dict , _A : List[str]=None ): _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + class_code if overwrite_result is not None: _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + overwrite_result _UpperCamelCase = black.Mode(target_versions={black.TargetVersion.PYaa} , line_length=119 ) _UpperCamelCase = black.format_str(_A , mode=_A ) _UpperCamelCase = os.path.join(self.diffusers_dir , '''new_code.py''' ) with open(_A , '''w''' , newline='''\n''' ) as f: f.write(_A ) if overwrite_result is None: self.assertTrue(len(check_copies.is_copy_consistent(_A ) ) == 0 ) else: check_copies.is_copy_consistent(f.name , overwrite=_A ) with open(_A , '''r''' ) as f: self.assertTrue(f.read() , _A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = check_copies.find_code_in_diffusers('''schedulers.scheduling_ddpm.DDPMSchedulerOutput''' ) self.assertEqual(_A , _A ) def UpperCamelCase_ ( self : Optional[Any] ): # Base copy consistency self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , REFERENCE_CODE + '''\n''' , ) # With no empty line at the end self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , _A , ) # Copy consistency with rename self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , re.sub('''DDPM''' , '''Test''' , _A ) , ) # Copy consistency with a really long name _UpperCamelCase = '''TestClassWithAReallyLongNameBecauseSomePeopleLikeThatForSomeReason''' self.check_copy_consistency( F"""# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->{long_class_name}""" , F"""{long_class_name}SchedulerOutput""" , re.sub('''Bert''' , _A , _A ) , ) # Copy consistency with overwrite self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , _A , overwrite_result=re.sub('''DDPM''' , '''Test''' , _A ) , )
71
1
import argparse import glob import logging import os import sys import time from collections import defaultdict from pathlib import Path from typing import Dict, List, Tuple import numpy as np import pytorch_lightning as pl import torch from callbacks import SeqaSeqLoggingCallback, get_checkpoint_callback, get_early_stopping_callback from torch import nn from torch.utils.data import DataLoader from transformers import MBartTokenizer, TaForConditionalGeneration from transformers.models.bart.modeling_bart import shift_tokens_right from utils import ( ROUGE_KEYS, LegacySeqaSeqDataset, SeqaSeqDataset, assert_all_frozen, calculate_bleu, calculate_rouge, check_output_dir, flatten_list, freeze_embeds, freeze_params, get_git_info, label_smoothed_nll_loss, lmap, pickle_save, save_git_info, save_json, use_task_specific_params, ) # need the parent dir module sys.path.insert(2, str(Path(__file__).resolve().parents[1])) from lightning_base import BaseTransformer, add_generic_args, generic_train # noqa _lowerCAmelCase = logging.getLogger(__name__) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "summarization" UpperCAmelCase = ["loss"] UpperCAmelCase = ROUGE_KEYS UpperCAmelCase = "rouge2" def __init__( self : str , _A : int , **_A : Optional[int] ): if hparams.sortish_sampler and hparams.gpus > 1: _UpperCamelCase = False elif hparams.max_tokens_per_batch is not None: if hparams.gpus > 1: raise NotImplementedError('''Dynamic Batch size does not work for multi-gpu training''' ) if hparams.sortish_sampler: raise ValueError('''--sortish_sampler and --max_tokens_per_batch may not be used simultaneously''' ) super().__init__(_A , num_labels=_A , mode=self.mode , **_A ) use_task_specific_params(self.model , '''summarization''' ) save_git_info(self.hparams.output_dir ) _UpperCamelCase = Path(self.output_dir ) / '''metrics.json''' _UpperCamelCase = Path(self.output_dir ) / '''hparams.pkl''' pickle_save(self.hparams , self.hparams_save_path ) _UpperCamelCase = 0 _UpperCamelCase = defaultdict(_A ) _UpperCamelCase = self.config.model_type _UpperCamelCase = self.config.tgt_vocab_size if self.model_type == '''fsmt''' else self.config.vocab_size _UpperCamelCase = { "data_dir": self.hparams.data_dir, "max_source_length": self.hparams.max_source_length, "prefix": self.model.config.prefix or "", } _UpperCamelCase = { '''train''': self.hparams.n_train, '''val''': self.hparams.n_val, '''test''': self.hparams.n_test, } _UpperCamelCase = {k: v if v >= 0 else None for k, v in n_observations_per_split.items()} _UpperCamelCase = { '''train''': self.hparams.max_target_length, '''val''': self.hparams.val_max_target_length, '''test''': self.hparams.test_max_target_length, } assert self.target_lens["train"] <= self.target_lens["val"], F"""target_lens: {self.target_lens}""" assert self.target_lens["train"] <= self.target_lens["test"], F"""target_lens: {self.target_lens}""" if self.hparams.freeze_embeds: freeze_embeds(self.model ) if self.hparams.freeze_encoder: freeze_params(self.model.get_encoder() ) assert_all_frozen(self.model.get_encoder() ) _UpperCamelCase = get_git_info()['''repo_sha'''] _UpperCamelCase = hparams.num_workers _UpperCamelCase = None # default to config if self.model.config.decoder_start_token_id is None and isinstance(self.tokenizer , _A ): _UpperCamelCase = self.tokenizer.lang_code_to_id[hparams.tgt_lang] _UpperCamelCase = self.decoder_start_token_id _UpperCamelCase = ( SeqaSeqDataset if hasattr(self.tokenizer , '''prepare_seq2seq_batch''' ) else LegacySeqaSeqDataset ) _UpperCamelCase = False _UpperCamelCase = self.model.config.num_beams if self.hparams.eval_beams is None else self.hparams.eval_beams if self.hparams.eval_max_gen_length is not None: _UpperCamelCase = self.hparams.eval_max_gen_length else: _UpperCamelCase = self.model.config.max_length _UpperCamelCase = self.default_val_metric if self.hparams.val_metric is None else self.hparams.val_metric def UpperCamelCase_ ( self : str , _A : Dict[str, torch.Tensor] ): _UpperCamelCase = { k: self.tokenizer.batch_decode(v.tolist() ) if '''mask''' not in k else v.shape for k, v in batch.items() } save_json(_A , Path(self.output_dir ) / '''text_batch.json''' ) save_json({k: v.tolist() for k, v in batch.items()} , Path(self.output_dir ) / '''tok_batch.json''' ) _UpperCamelCase = True return readable_batch def UpperCamelCase_ ( self : Tuple , _A : Tuple , **_A : Dict ): return self.model(_A , **_A ) def UpperCamelCase_ ( self : Any , _A : List[int] ): _UpperCamelCase = self.tokenizer.batch_decode( _A , skip_special_tokens=_A , clean_up_tokenization_spaces=_A ) return lmap(str.strip , _A ) def UpperCamelCase_ ( self : List[str] , _A : dict ): _UpperCamelCase = self.tokenizer.pad_token_id _UpperCamelCase , _UpperCamelCase = batch['''input_ids'''], batch['''attention_mask'''] _UpperCamelCase = batch['''labels'''] if isinstance(self.model , _A ): _UpperCamelCase = self.model._shift_right(_A ) else: _UpperCamelCase = shift_tokens_right(_A , _A ) if not self.already_saved_batch: # This would be slightly better if it only happened on rank zero _UpperCamelCase = decoder_input_ids self.save_readable_batch(_A ) _UpperCamelCase = self(_A , attention_mask=_A , decoder_input_ids=_A , use_cache=_A ) _UpperCamelCase = outputs['''logits'''] if self.hparams.label_smoothing == 0: # Same behavior as modeling_bart.py, besides ignoring pad_token_id _UpperCamelCase = nn.CrossEntropyLoss(ignore_index=_A ) assert lm_logits.shape[-1] == self.vocab_size _UpperCamelCase = ce_loss_fct(lm_logits.view(-1 , lm_logits.shape[-1] ) , tgt_ids.view(-1 ) ) else: _UpperCamelCase = nn.functional.log_softmax(_A , dim=-1 ) _UpperCamelCase , _UpperCamelCase = label_smoothed_nll_loss( _A , _A , self.hparams.label_smoothing , ignore_index=_A ) return (loss,) @property def UpperCamelCase_ ( self : Optional[Any] ): return self.tokenizer.pad_token_id def UpperCamelCase_ ( self : Optional[int] , _A : str , _A : Optional[Any] ): _UpperCamelCase = self._step(_A ) _UpperCamelCase = dict(zip(self.loss_names , _A ) ) # tokens per batch _UpperCamelCase = batch['''input_ids'''].ne(self.pad ).sum() + batch['''labels'''].ne(self.pad ).sum() _UpperCamelCase = batch['''input_ids'''].shape[0] _UpperCamelCase = batch['''input_ids'''].eq(self.pad ).sum() _UpperCamelCase = batch['''input_ids'''].eq(self.pad ).float().mean() # TODO(SS): make a wandb summary metric for this return {"loss": loss_tensors[0], "log": logs} def UpperCamelCase_ ( self : str , _A : List[str] , _A : Tuple ): return self._generative_step(_A ) def UpperCamelCase_ ( self : Dict , _A : Optional[Any] , _A : Optional[int]="val" ): self.step_count += 1 _UpperCamelCase = {k: torch.stack([x[k] for x in outputs] ).mean() for k in self.loss_names} _UpperCamelCase = losses['''loss'''] _UpperCamelCase = { k: np.array([x[k] for x in outputs] ).mean() for k in self.metric_names + ['''gen_time''', '''gen_len'''] } _UpperCamelCase = ( generative_metrics[self.val_metric] if self.val_metric in generative_metrics else losses[self.val_metric] ) _UpperCamelCase = torch.tensor(_A ).type_as(_A ) generative_metrics.update({k: v.item() for k, v in losses.items()} ) losses.update(_A ) _UpperCamelCase = {F"""{prefix}_avg_{k}""": x for k, x in losses.items()} _UpperCamelCase = self.step_count self.metrics[prefix].append(_A ) # callback writes this to self.metrics_save_path _UpperCamelCase = flatten_list([x['''preds'''] for x in outputs] ) return { "log": all_metrics, "preds": preds, F"""{prefix}_loss""": loss, F"""{prefix}_{self.val_metric}""": metric_tensor, } def UpperCamelCase_ ( self : Any , _A : str , _A : Dict ): return calculate_rouge(_A , _A ) def UpperCamelCase_ ( self : Tuple , _A : dict ): _UpperCamelCase = time.time() # parser.add_argument('--eval_max_gen_length', type=int, default=None, help='never generate more than n tokens') _UpperCamelCase = self.model.generate( batch['''input_ids'''] , attention_mask=batch['''attention_mask'''] , use_cache=_A , decoder_start_token_id=self.decoder_start_token_id , num_beams=self.eval_beams , max_length=self.eval_max_length , ) _UpperCamelCase = (time.time() - ta) / batch['''input_ids'''].shape[0] _UpperCamelCase = self.ids_to_clean_text(_A ) _UpperCamelCase = self.ids_to_clean_text(batch['''labels'''] ) _UpperCamelCase = self._step(_A ) _UpperCamelCase = dict(zip(self.loss_names , _A ) ) _UpperCamelCase = self.calc_generative_metrics(_A , _A ) _UpperCamelCase = np.mean(lmap(_A , _A ) ) base_metrics.update(gen_time=_A , gen_len=_A , preds=_A , target=_A , **_A ) return base_metrics def UpperCamelCase_ ( self : int , _A : Tuple , _A : Tuple ): return self._generative_step(_A ) def UpperCamelCase_ ( self : Any , _A : List[Any] ): return self.validation_epoch_end(_A , prefix='''test''' ) def UpperCamelCase_ ( self : Any , _A : Union[str, Any] ): _UpperCamelCase = self.n_obs[type_path] _UpperCamelCase = self.target_lens[type_path] _UpperCamelCase = self.dataset_class( self.tokenizer , type_path=_A , n_obs=_A , max_target_length=_A , **self.dataset_kwargs , ) return dataset def UpperCamelCase_ ( self : Any , _A : str , _A : int , _A : bool = False ): _UpperCamelCase = self.get_dataset(_A ) if self.hparams.sortish_sampler and type_path != "test" and type_path != "val": _UpperCamelCase = dataset.make_sortish_sampler(_A , distributed=self.hparams.gpus > 1 ) return DataLoader( _A , batch_size=_A , collate_fn=dataset.collate_fn , shuffle=_A , num_workers=self.num_workers , sampler=_A , ) elif self.hparams.max_tokens_per_batch is not None and type_path != "test" and type_path != "val": _UpperCamelCase = dataset.make_dynamic_sampler( self.hparams.max_tokens_per_batch , distributed=self.hparams.gpus > 1 ) return DataLoader( _A , batch_sampler=_A , collate_fn=dataset.collate_fn , num_workers=self.num_workers , ) else: return DataLoader( _A , batch_size=_A , collate_fn=dataset.collate_fn , shuffle=_A , num_workers=self.num_workers , sampler=_A , ) def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = self.get_dataloader('''train''' , batch_size=self.hparams.train_batch_size , shuffle=_A ) return dataloader def UpperCamelCase_ ( self : int ): return self.get_dataloader('''val''' , batch_size=self.hparams.eval_batch_size ) def UpperCamelCase_ ( self : Optional[int] ): return self.get_dataloader('''test''' , batch_size=self.hparams.eval_batch_size ) @staticmethod def UpperCamelCase_ ( _A : int , _A : List[Any] ): BaseTransformer.add_model_specific_args(_A , _A ) add_generic_args(_A , _A ) parser.add_argument( '''--max_source_length''' , default=1024 , type=_A , help=( '''The maximum total input sequence length after tokenization. Sequences longer ''' '''than this will be truncated, sequences shorter will be padded.''' ) , ) parser.add_argument( '''--max_target_length''' , default=56 , type=_A , help=( '''The maximum total input sequence length after tokenization. Sequences longer ''' '''than this will be truncated, sequences shorter will be padded.''' ) , ) parser.add_argument( '''--val_max_target_length''' , default=142 , type=_A , help=( '''The maximum total input sequence length after tokenization. Sequences longer ''' '''than this will be truncated, sequences shorter will be padded.''' ) , ) parser.add_argument( '''--test_max_target_length''' , default=142 , type=_A , help=( '''The maximum total input sequence length after tokenization. Sequences longer ''' '''than this will be truncated, sequences shorter will be padded.''' ) , ) parser.add_argument('''--freeze_encoder''' , action='''store_true''' ) parser.add_argument('''--freeze_embeds''' , action='''store_true''' ) parser.add_argument('''--sortish_sampler''' , action='''store_true''' , default=_A ) parser.add_argument('''--overwrite_output_dir''' , action='''store_true''' , default=_A ) parser.add_argument('''--max_tokens_per_batch''' , type=_A , default=_A ) parser.add_argument('''--logger_name''' , type=_A , choices=['''default''', '''wandb''', '''wandb_shared'''] , default='''default''' ) parser.add_argument('''--n_train''' , type=_A , default=-1 , required=_A , help='''# examples. -1 means use all.''' ) parser.add_argument('''--n_val''' , type=_A , default=500 , required=_A , help='''# examples. -1 means use all.''' ) parser.add_argument('''--n_test''' , type=_A , default=-1 , required=_A , help='''# examples. -1 means use all.''' ) parser.add_argument( '''--task''' , type=_A , default='''summarization''' , required=_A , help='''# examples. -1 means use all.''' ) parser.add_argument('''--label_smoothing''' , type=_A , default=0.0 , required=_A ) parser.add_argument('''--src_lang''' , type=_A , default='''''' , required=_A ) parser.add_argument('''--tgt_lang''' , type=_A , default='''''' , required=_A ) parser.add_argument('''--eval_beams''' , type=_A , default=_A , required=_A ) parser.add_argument( '''--val_metric''' , type=_A , default=_A , required=_A , choices=['''bleu''', '''rouge2''', '''loss''', None] ) parser.add_argument('''--eval_max_gen_length''' , type=_A , default=_A , help='''never generate more than n tokens''' ) parser.add_argument('''--save_top_k''' , type=_A , default=1 , required=_A , help='''How many checkpoints to save''' ) parser.add_argument( '''--early_stopping_patience''' , type=_A , default=-1 , required=_A , help=( '''-1 means never early stop. early_stopping_patience is measured in validation checks, not epochs. So''' ''' val_check_interval will effect it.''' ) , ) return parser class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "translation" UpperCAmelCase = ["loss"] UpperCAmelCase = ["bleu"] UpperCAmelCase = "bleu" def __init__( self : Union[str, Any] , _A : Tuple , **_A : List[str] ): super().__init__(_A , **_A ) _UpperCamelCase = hparams.src_lang _UpperCamelCase = hparams.tgt_lang def UpperCamelCase_ ( self : List[Any] , _A : Dict , _A : List[str] ): return calculate_bleu(_A , _A ) def _snake_case ( __snake_case , __snake_case=None ): Path(args.output_dir ).mkdir(exist_ok=__snake_case ) check_output_dir(__snake_case , expected_items=3 ) if model is None: if "summarization" in args.task: _UpperCamelCase = SummarizationModule(__snake_case ) else: _UpperCamelCase = TranslationModule(__snake_case ) _UpperCamelCase = Path(args.data_dir ).name if ( args.logger_name == "default" or args.fast_dev_run or str(args.output_dir ).startswith('''/tmp''' ) or str(args.output_dir ).startswith('''/var''' ) ): _UpperCamelCase = True # don't pollute wandb logs unnecessarily elif args.logger_name == "wandb": from pytorch_lightning.loggers import WandbLogger _UpperCamelCase = os.environ.get('''WANDB_PROJECT''' , __snake_case ) _UpperCamelCase = WandbLogger(name=model.output_dir.name , project=__snake_case ) elif args.logger_name == "wandb_shared": from pytorch_lightning.loggers import WandbLogger _UpperCamelCase = WandbLogger(name=model.output_dir.name , project=f"""hf_{dataset}""" ) if args.early_stopping_patience >= 0: _UpperCamelCase = get_early_stopping_callback(model.val_metric , args.early_stopping_patience ) else: _UpperCamelCase = False _UpperCamelCase = args.val_metric == '''loss''' _UpperCamelCase = generic_train( __snake_case , __snake_case , logging_callback=SeqaSeqLoggingCallback() , checkpoint_callback=get_checkpoint_callback( args.output_dir , model.val_metric , args.save_top_k , __snake_case ) , early_stopping_callback=__snake_case , logger=__snake_case , ) pickle_save(model.hparams , model.output_dir / '''hparams.pkl''' ) if not args.do_predict: return model _UpperCamelCase = '''''' _UpperCamelCase = sorted(glob.glob(os.path.join(args.output_dir , '''*.ckpt''' ) , recursive=__snake_case ) ) if checkpoints: _UpperCamelCase = checkpoints[-1] _UpperCamelCase = checkpoints[-1] trainer.logger.log_hyperparams(model.hparams ) # test() without a model tests using the best checkpoint automatically trainer.test() return model if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() _lowerCAmelCase = pl.Trainer.add_argparse_args(parser) _lowerCAmelCase = SummarizationModule.add_model_specific_args(parser, os.getcwd()) _lowerCAmelCase = parser.parse_args() main(args)
71
from __future__ import annotations import math class lowerCAmelCase_ : def __init__( self : int , _A : int ): _UpperCamelCase = size # approximate the overall size of segment tree with given value _UpperCamelCase = [0 for i in range(0 , 4 * size )] # create array to store lazy update _UpperCamelCase = [0 for i in range(0 , 4 * size )] _UpperCamelCase = [0 for i in range(0 , 4 * size )] # flag for lazy update def UpperCamelCase_ ( self : str , _A : int ): return idx * 2 def UpperCamelCase_ ( self : Any , _A : int ): return idx * 2 + 1 def UpperCamelCase_ ( self : Union[str, Any] , _A : int , _A : int , _A : int , _A : list[int] ): if left_element == right_element: _UpperCamelCase = a[left_element - 1] else: _UpperCamelCase = (left_element + right_element) // 2 self.build(self.left(_A ) , _A , _A , _A ) self.build(self.right(_A ) , mid + 1 , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) def UpperCamelCase_ ( self : Tuple , _A : int , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return True if left_element >= a and right_element <= b: _UpperCamelCase = val if left_element != right_element: _UpperCamelCase = val _UpperCamelCase = val _UpperCamelCase = True _UpperCamelCase = True return True _UpperCamelCase = (left_element + right_element) // 2 self.update(self.left(_A ) , _A , _A , _A , _A , _A ) self.update(self.right(_A ) , mid + 1 , _A , _A , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) return True def UpperCamelCase_ ( self : Any , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return -math.inf if left_element >= a and right_element <= b: return self.segment_tree[idx] _UpperCamelCase = (left_element + right_element) // 2 _UpperCamelCase = self.query(self.left(_A ) , _A , _A , _A , _A ) _UpperCamelCase = self.query(self.right(_A ) , mid + 1 , _A , _A , _A ) return max(_A , _A ) def __str__( self : Tuple ): return str([self.query(1 , 1 , self.size , _A , _A ) for i in range(1 , self.size + 1 )] ) if __name__ == "__main__": _lowerCAmelCase = [1, 2, -4, 7, 3, -5, 6, 11, -20, 9, 14, 15, 5, 2, -8] _lowerCAmelCase = 15 _lowerCAmelCase = SegmentTree(size) segt.build(1, 1, size, A) print(segt.query(1, 1, size, 4, 6)) print(segt.query(1, 1, size, 7, 11)) print(segt.query(1, 1, size, 7, 12)) segt.update(1, 1, size, 1, 3, 111) print(segt.query(1, 1, size, 1, 15)) segt.update(1, 1, size, 7, 8, 235) print(segt)
71
1
import copy from collections import OrderedDict from typing import Mapping from packaging import version from ...configuration_utils import PretrainedConfig from ...onnx import OnnxConfig from ...utils import logging from ..auto import CONFIG_MAPPING _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "microsoft/conditional-detr-resnet-50": ( "https://huggingface.co/microsoft/conditional-detr-resnet-50/resolve/main/config.json" ), } class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "conditional_detr" UpperCAmelCase = ["past_key_values"] UpperCAmelCase = { "hidden_size": "d_model", "num_attention_heads": "encoder_attention_heads", } def __init__( self : List[Any] , _A : str=True , _A : str=None , _A : Optional[int]=3 , _A : Dict=300 , _A : Union[str, Any]=6 , _A : Any=2048 , _A : Dict=8 , _A : List[str]=6 , _A : List[Any]=2048 , _A : Tuple=8 , _A : int=0.0 , _A : List[str]=0.0 , _A : str=True , _A : Any="relu" , _A : Dict=256 , _A : List[str]=0.1 , _A : Optional[int]=0.0 , _A : List[str]=0.0 , _A : str=0.02 , _A : Tuple=1.0 , _A : Optional[Any]=False , _A : int="sine" , _A : Optional[Any]="resnet50" , _A : str=True , _A : str=False , _A : Optional[int]=2 , _A : Union[str, Any]=5 , _A : str=2 , _A : Optional[Any]=1 , _A : str=1 , _A : List[str]=2 , _A : List[str]=5 , _A : str=2 , _A : Dict=0.25 , **_A : str , ): if backbone_config is not None and use_timm_backbone: raise ValueError('''You can\'t specify both `backbone_config` and `use_timm_backbone`.''' ) if not use_timm_backbone: if backbone_config is None: logger.info('''`backbone_config` is `None`. Initializing the config with the default `ResNet` backbone.''' ) _UpperCamelCase = CONFIG_MAPPING['''resnet'''](out_features=['''stage4'''] ) elif isinstance(_A , _A ): _UpperCamelCase = backbone_config.get('''model_type''' ) _UpperCamelCase = CONFIG_MAPPING[backbone_model_type] _UpperCamelCase = config_class.from_dict(_A ) _UpperCamelCase = use_timm_backbone _UpperCamelCase = backbone_config _UpperCamelCase = num_channels _UpperCamelCase = num_queries _UpperCamelCase = d_model _UpperCamelCase = encoder_ffn_dim _UpperCamelCase = encoder_layers _UpperCamelCase = encoder_attention_heads _UpperCamelCase = decoder_ffn_dim _UpperCamelCase = decoder_layers _UpperCamelCase = decoder_attention_heads _UpperCamelCase = dropout _UpperCamelCase = attention_dropout _UpperCamelCase = activation_dropout _UpperCamelCase = activation_function _UpperCamelCase = init_std _UpperCamelCase = init_xavier_std _UpperCamelCase = encoder_layerdrop _UpperCamelCase = decoder_layerdrop _UpperCamelCase = encoder_layers _UpperCamelCase = auxiliary_loss _UpperCamelCase = position_embedding_type _UpperCamelCase = backbone _UpperCamelCase = use_pretrained_backbone _UpperCamelCase = dilation # Hungarian matcher _UpperCamelCase = class_cost _UpperCamelCase = bbox_cost _UpperCamelCase = giou_cost # Loss coefficients _UpperCamelCase = mask_loss_coefficient _UpperCamelCase = dice_loss_coefficient _UpperCamelCase = cls_loss_coefficient _UpperCamelCase = bbox_loss_coefficient _UpperCamelCase = giou_loss_coefficient _UpperCamelCase = focal_alpha super().__init__(is_encoder_decoder=_A , **_A ) @property def UpperCamelCase_ ( self : Optional[Any] ): return self.encoder_attention_heads @property def UpperCamelCase_ ( self : Tuple ): return self.d_model def UpperCamelCase_ ( self : str ): _UpperCamelCase = copy.deepcopy(self.__dict__ ) if self.backbone_config is not None: _UpperCamelCase = self.backbone_config.to_dict() _UpperCamelCase = self.__class__.model_type return output class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = version.parse("1.11" ) @property def UpperCamelCase_ ( self : str ): return OrderedDict( [ ('''pixel_values''', {0: '''batch''', 1: '''num_channels''', 2: '''height''', 3: '''width'''}), ('''pixel_mask''', {0: '''batch'''}), ] ) @property def UpperCamelCase_ ( self : int ): return 1e-5 @property def UpperCamelCase_ ( self : Any ): return 12
71
from typing import TYPE_CHECKING from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available _lowerCAmelCase = { "configuration_jukebox": [ "JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP", "JukeboxConfig", "JukeboxPriorConfig", "JukeboxVQVAEConfig", ], "tokenization_jukebox": ["JukeboxTokenizer"], } try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST", "JukeboxModel", "JukeboxPreTrainedModel", "JukeboxVQVAE", "JukeboxPrior", ] if TYPE_CHECKING: from .configuration_jukebox import ( JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP, JukeboxConfig, JukeboxPriorConfig, JukeboxVQVAEConfig, ) from .tokenization_jukebox import JukeboxTokenizer try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_jukebox import ( JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST, JukeboxModel, JukeboxPreTrainedModel, JukeboxPrior, JukeboxVQVAE, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
1
from collections.abc import Generator from math import sin def _snake_case ( __snake_case ): if len(__snake_case ) != 32: raise ValueError('''Input must be of length 32''' ) _UpperCamelCase = B'''''' for i in [3, 2, 1, 0]: little_endian += string_aa[8 * i : 8 * i + 8] return little_endian def _snake_case ( __snake_case ): if i < 0: raise ValueError('''Input must be non-negative''' ) _UpperCamelCase = format(__snake_case , '''08x''' )[-8:] _UpperCamelCase = B'''''' for i in [3, 2, 1, 0]: little_endian_hex += hex_rep[2 * i : 2 * i + 2].encode('''utf-8''' ) return little_endian_hex def _snake_case ( __snake_case ): _UpperCamelCase = B'''''' for char in message: bit_string += format(__snake_case , '''08b''' ).encode('''utf-8''' ) _UpperCamelCase = format(len(__snake_case ) , '''064b''' ).encode('''utf-8''' ) # Pad bit_string to a multiple of 512 chars bit_string += b"1" while len(__snake_case ) % 512 != 448: bit_string += b"0" bit_string += to_little_endian(start_len[32:] ) + to_little_endian(start_len[:32] ) return bit_string def _snake_case ( __snake_case ): if len(__snake_case ) % 512 != 0: raise ValueError('''Input must have length that\'s a multiple of 512''' ) for pos in range(0 , len(__snake_case ) , 512 ): _UpperCamelCase = bit_string[pos : pos + 512] _UpperCamelCase = [] for i in range(0 , 512 , 32 ): block_words.append(int(to_little_endian(block[i : i + 32] ) , 2 ) ) yield block_words def _snake_case ( __snake_case ): if i < 0: raise ValueError('''Input must be non-negative''' ) _UpperCamelCase = format(__snake_case , '''032b''' ) _UpperCamelCase = '''''' for c in i_str: new_str += "1" if c == "0" else "0" return int(__snake_case , 2 ) def _snake_case ( __snake_case , __snake_case ): return (a + b) % 2**32 def _snake_case ( __snake_case , __snake_case ): if i < 0: raise ValueError('''Input must be non-negative''' ) if shift < 0: raise ValueError('''Shift must be non-negative''' ) return ((i << shift) ^ (i >> (32 - shift))) % 2**32 def _snake_case ( __snake_case ): _UpperCamelCase = preprocess(__snake_case ) _UpperCamelCase = [int(2**32 * abs(sin(i + 1 ) ) ) for i in range(64 )] # Starting states _UpperCamelCase = 0x6745_2301 _UpperCamelCase = 0xEFCD_AB89 _UpperCamelCase = 0x98BA_DCFE _UpperCamelCase = 0x1032_5476 _UpperCamelCase = [ 7, 12, 17, 22, 7, 12, 17, 22, 7, 12, 17, 22, 7, 12, 17, 22, 5, 9, 14, 20, 5, 9, 14, 20, 5, 9, 14, 20, 5, 9, 14, 20, 4, 11, 16, 23, 4, 11, 16, 23, 4, 11, 16, 23, 4, 11, 16, 23, 6, 10, 15, 21, 6, 10, 15, 21, 6, 10, 15, 21, 6, 10, 15, 21, ] # Process bit string in chunks, each with 16 32-char words for block_words in get_block_words(__snake_case ): _UpperCamelCase = aa _UpperCamelCase = ba _UpperCamelCase = ca _UpperCamelCase = da # Hash current chunk for i in range(64 ): if i <= 15: # f = (b & c) | (not_32(b) & d) # Alternate definition for f _UpperCamelCase = d ^ (b & (c ^ d)) _UpperCamelCase = i elif i <= 31: # f = (d & b) | (not_32(d) & c) # Alternate definition for f _UpperCamelCase = c ^ (d & (b ^ c)) _UpperCamelCase = (5 * i + 1) % 16 elif i <= 47: _UpperCamelCase = b ^ c ^ d _UpperCamelCase = (3 * i + 5) % 16 else: _UpperCamelCase = c ^ (b | not_aa(__snake_case )) _UpperCamelCase = (7 * i) % 16 _UpperCamelCase = (f + a + added_consts[i] + block_words[g]) % 2**32 _UpperCamelCase = d _UpperCamelCase = c _UpperCamelCase = b _UpperCamelCase = sum_aa(__snake_case , left_rotate_aa(__snake_case , shift_amounts[i] ) ) # Add hashed chunk to running total _UpperCamelCase = sum_aa(__snake_case , __snake_case ) _UpperCamelCase = sum_aa(__snake_case , __snake_case ) _UpperCamelCase = sum_aa(__snake_case , __snake_case ) _UpperCamelCase = sum_aa(__snake_case , __snake_case ) _UpperCamelCase = reformat_hex(__snake_case ) + reformat_hex(__snake_case ) + reformat_hex(__snake_case ) + reformat_hex(__snake_case ) return digest if __name__ == "__main__": import doctest doctest.testmod()
71
import multiprocessing import os from typing import BinaryIO, Optional, Union import fsspec from .. import Dataset, Features, NamedSplit, config from ..formatting import query_table from ..packaged_modules.json.json import Json from ..utils import logging from ..utils.typing import NestedDataStructureLike, PathLike from .abc import AbstractDatasetReader class lowerCAmelCase_ ( __lowercase ): def __init__( self : int , _A : NestedDataStructureLike[PathLike] , _A : Optional[NamedSplit] = None , _A : Optional[Features] = None , _A : str = None , _A : bool = False , _A : bool = False , _A : Optional[str] = None , _A : Optional[int] = None , **_A : str , ): super().__init__( _A , split=_A , features=_A , cache_dir=_A , keep_in_memory=_A , streaming=_A , num_proc=_A , **_A , ) _UpperCamelCase = field _UpperCamelCase = path_or_paths if isinstance(_A , _A ) else {self.split: path_or_paths} _UpperCamelCase = Json( cache_dir=_A , data_files=_A , features=_A , field=_A , **_A , ) def UpperCamelCase_ ( self : List[str] ): # Build iterable dataset if self.streaming: _UpperCamelCase = self.builder.as_streaming_dataset(split=self.split ) # Build regular (map-style) dataset else: _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None self.builder.download_and_prepare( download_config=_A , download_mode=_A , verification_mode=_A , base_path=_A , num_proc=self.num_proc , ) _UpperCamelCase = self.builder.as_dataset( split=self.split , verification_mode=_A , in_memory=self.keep_in_memory ) return dataset class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : Dataset , _A : Union[PathLike, BinaryIO] , _A : Optional[int] = None , _A : Optional[int] = None , **_A : List[str] , ): if num_proc is not None and num_proc <= 0: raise ValueError(F"""num_proc {num_proc} must be an integer > 0.""" ) _UpperCamelCase = dataset _UpperCamelCase = path_or_buf _UpperCamelCase = batch_size if batch_size else config.DEFAULT_MAX_BATCH_SIZE _UpperCamelCase = num_proc _UpperCamelCase = '''utf-8''' _UpperCamelCase = to_json_kwargs def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = self.to_json_kwargs.pop('''path_or_buf''' , _A ) _UpperCamelCase = self.to_json_kwargs.pop('''orient''' , '''records''' ) _UpperCamelCase = self.to_json_kwargs.pop('''lines''' , True if orient == '''records''' else False ) _UpperCamelCase = self.to_json_kwargs.pop('''index''' , False if orient in ['''split''', '''table'''] else True ) _UpperCamelCase = self.to_json_kwargs.pop('''compression''' , _A ) if compression not in [None, "infer", "gzip", "bz2", "xz"]: raise NotImplementedError(F"""`datasets` currently does not support {compression} compression""" ) if isinstance(self.path_or_buf , (str, bytes, os.PathLike) ): with fsspec.open(self.path_or_buf , '''wb''' , compression=_A ) as buffer: _UpperCamelCase = self._write(file_obj=_A , orient=_A , lines=_A , index=_A , **self.to_json_kwargs ) else: if compression: raise NotImplementedError( F"""The compression parameter is not supported when writing to a buffer, but compression={compression}""" ''' was passed. Please provide a local path instead.''' ) _UpperCamelCase = self._write( file_obj=self.path_or_buf , orient=_A , lines=_A , index=_A , **self.to_json_kwargs ) return written def UpperCamelCase_ ( self : Any , _A : Optional[Any] ): _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = args _UpperCamelCase = query_table( table=self.dataset.data , key=slice(_A , offset + self.batch_size ) , indices=self.dataset._indices , ) _UpperCamelCase = batch.to_pandas().to_json( path_or_buf=_A , orient=_A , lines=_A , index=_A , **_A ) if not json_str.endswith('''\n''' ): json_str += "\n" return json_str.encode(self.encoding ) def UpperCamelCase_ ( self : int , _A : BinaryIO , _A : Dict , _A : Optional[Any] , _A : Dict , **_A : str , ): _UpperCamelCase = 0 if self.num_proc is None or self.num_proc == 1: for offset in logging.tqdm( range(0 , len(self.dataset ) , self.batch_size ) , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating json from Arrow format''' , ): _UpperCamelCase = self._batch_json((offset, orient, lines, index, to_json_kwargs) ) written += file_obj.write(_A ) else: _UpperCamelCase , _UpperCamelCase = len(self.dataset ), self.batch_size with multiprocessing.Pool(self.num_proc ) as pool: for json_str in logging.tqdm( pool.imap( self._batch_json , [(offset, orient, lines, index, to_json_kwargs) for offset in range(0 , _A , _A )] , ) , total=(num_rows // batch_size) + 1 if num_rows % batch_size else num_rows // batch_size , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating json from Arrow format''' , ): written += file_obj.write(_A ) return written
71
1
import collections import os import re from pathlib import Path _lowerCAmelCase = "src/transformers" # Matches is_xxx_available() _lowerCAmelCase = re.compile(r"is\_([a-z_]*)_available()") # Catches a one-line _import_struct = {xxx} _lowerCAmelCase = re.compile(r"^_import_structure\s+=\s+\{([^\}]+)\}") # Catches a line with a key-values pattern: "bla": ["foo", "bar"] _lowerCAmelCase = re.compile(r"\s+\"\S*\":\s+\[([^\]]*)\]") # Catches a line if not is_foo_available _lowerCAmelCase = re.compile(r"^\s*if\s+not\s+is\_[a-z_]*\_available\(\)") # Catches a line _import_struct["bla"].append("foo") _lowerCAmelCase = re.compile(r"^\s*_import_structure\[\"\S*\"\]\.append\(\"(\S*)\"\)") # Catches a line _import_struct["bla"].extend(["foo", "bar"]) or _import_struct["bla"] = ["foo", "bar"] _lowerCAmelCase = re.compile(r"^\s*_import_structure\[\S*\](?:\.extend\(|\s*=\s+)\[([^\]]*)\]") # Catches a line with an object between quotes and a comma: "MyModel", _lowerCAmelCase = re.compile(r"^\s+\"([^\"]+)\",") # Catches a line with objects between brackets only: ["foo", "bar"], _lowerCAmelCase = re.compile(r"^\s+\[([^\]]+)\]") # Catches a line with from foo import bar, bla, boo _lowerCAmelCase = re.compile(r"\s+from\s+\S*\s+import\s+([^\(\s].*)\n") # Catches a line with try: _lowerCAmelCase = re.compile(r"^\s*try:") # Catches a line with else: _lowerCAmelCase = re.compile(r"^\s*else:") def _snake_case ( __snake_case ): if _re_test_backend.search(__snake_case ) is None: return None _UpperCamelCase = [b[0] for b in _re_backend.findall(__snake_case )] backends.sort() return "_and_".join(__snake_case ) def _snake_case ( __snake_case ): with open(__snake_case , '''r''' , encoding='''utf-8''' , newline='''\n''' ) as f: _UpperCamelCase = f.readlines() _UpperCamelCase = 0 while line_index < len(__snake_case ) and not lines[line_index].startswith('''_import_structure = {''' ): line_index += 1 # If this is a traditional init, just return. if line_index >= len(__snake_case ): return None # First grab the objects without a specific backend in _import_structure _UpperCamelCase = [] while not lines[line_index].startswith('''if TYPE_CHECKING''' ) and find_backend(lines[line_index] ) is None: _UpperCamelCase = lines[line_index] # If we have everything on a single line, let's deal with it. if _re_one_line_import_struct.search(__snake_case ): _UpperCamelCase = _re_one_line_import_struct.search(__snake_case ).groups()[0] _UpperCamelCase = re.findall(R'''\[([^\]]+)\]''' , __snake_case ) for imp in imports: objects.extend([obj[1:-1] for obj in imp.split(''', ''' )] ) line_index += 1 continue _UpperCamelCase = _re_import_struct_key_value.search(__snake_case ) if single_line_import_search is not None: _UpperCamelCase = [obj[1:-1] for obj in single_line_import_search.groups()[0].split(''', ''' ) if len(__snake_case ) > 0] objects.extend(__snake_case ) elif line.startswith(''' ''' * 8 + '''"''' ): objects.append(line[9:-3] ) line_index += 1 _UpperCamelCase = {'''none''': objects} # Let's continue with backend-specific objects in _import_structure while not lines[line_index].startswith('''if TYPE_CHECKING''' ): # If the line is an if not is_backend_available, we grab all objects associated. _UpperCamelCase = find_backend(lines[line_index] ) # Check if the backend declaration is inside a try block: if _re_try.search(lines[line_index - 1] ) is None: _UpperCamelCase = None if backend is not None: line_index += 1 # Scroll until we hit the else block of try-except-else while _re_else.search(lines[line_index] ) is None: line_index += 1 line_index += 1 _UpperCamelCase = [] # Until we unindent, add backend objects to the list while len(lines[line_index] ) <= 1 or lines[line_index].startswith(''' ''' * 4 ): _UpperCamelCase = lines[line_index] if _re_import_struct_add_one.search(__snake_case ) is not None: objects.append(_re_import_struct_add_one.search(__snake_case ).groups()[0] ) elif _re_import_struct_add_many.search(__snake_case ) is not None: _UpperCamelCase = _re_import_struct_add_many.search(__snake_case ).groups()[0].split(''', ''' ) _UpperCamelCase = [obj[1:-1] for obj in imports if len(__snake_case ) > 0] objects.extend(__snake_case ) elif _re_between_brackets.search(__snake_case ) is not None: _UpperCamelCase = _re_between_brackets.search(__snake_case ).groups()[0].split(''', ''' ) _UpperCamelCase = [obj[1:-1] for obj in imports if len(__snake_case ) > 0] objects.extend(__snake_case ) elif _re_quote_object.search(__snake_case ) is not None: objects.append(_re_quote_object.search(__snake_case ).groups()[0] ) elif line.startswith(''' ''' * 8 + '''"''' ): objects.append(line[9:-3] ) elif line.startswith(''' ''' * 12 + '''"''' ): objects.append(line[13:-3] ) line_index += 1 _UpperCamelCase = objects else: line_index += 1 # At this stage we are in the TYPE_CHECKING part, first grab the objects without a specific backend _UpperCamelCase = [] while ( line_index < len(__snake_case ) and find_backend(lines[line_index] ) is None and not lines[line_index].startswith('''else''' ) ): _UpperCamelCase = lines[line_index] _UpperCamelCase = _re_import.search(__snake_case ) if single_line_import_search is not None: objects.extend(single_line_import_search.groups()[0].split(''', ''' ) ) elif line.startswith(''' ''' * 8 ): objects.append(line[8:-2] ) line_index += 1 _UpperCamelCase = {'''none''': objects} # Let's continue with backend-specific objects while line_index < len(__snake_case ): # If the line is an if is_backend_available, we grab all objects associated. _UpperCamelCase = find_backend(lines[line_index] ) # Check if the backend declaration is inside a try block: if _re_try.search(lines[line_index - 1] ) is None: _UpperCamelCase = None if backend is not None: line_index += 1 # Scroll until we hit the else block of try-except-else while _re_else.search(lines[line_index] ) is None: line_index += 1 line_index += 1 _UpperCamelCase = [] # Until we unindent, add backend objects to the list while len(lines[line_index] ) <= 1 or lines[line_index].startswith(''' ''' * 8 ): _UpperCamelCase = lines[line_index] _UpperCamelCase = _re_import.search(__snake_case ) if single_line_import_search is not None: objects.extend(single_line_import_search.groups()[0].split(''', ''' ) ) elif line.startswith(''' ''' * 12 ): objects.append(line[12:-2] ) line_index += 1 _UpperCamelCase = objects else: line_index += 1 return import_dict_objects, type_hint_objects def _snake_case ( __snake_case , __snake_case ): def find_duplicates(__snake_case ): return [k for k, v in collections.Counter(__snake_case ).items() if v > 1] if list(import_dict_objects.keys() ) != list(type_hint_objects.keys() ): return ["Both sides of the init do not have the same backends!"] _UpperCamelCase = [] for key in import_dict_objects.keys(): _UpperCamelCase = find_duplicates(import_dict_objects[key] ) if duplicate_imports: errors.append(f"""Duplicate _import_structure definitions for: {duplicate_imports}""" ) _UpperCamelCase = find_duplicates(type_hint_objects[key] ) if duplicate_type_hints: errors.append(f"""Duplicate TYPE_CHECKING objects for: {duplicate_type_hints}""" ) if sorted(set(import_dict_objects[key] ) ) != sorted(set(type_hint_objects[key] ) ): _UpperCamelCase = '''base imports''' if key == '''none''' else f"""{key} backend""" errors.append(f"""Differences for {name}:""" ) for a in type_hint_objects[key]: if a not in import_dict_objects[key]: errors.append(f""" {a} in TYPE_HINT but not in _import_structure.""" ) for a in import_dict_objects[key]: if a not in type_hint_objects[key]: errors.append(f""" {a} in _import_structure but not in TYPE_HINT.""" ) return errors def _snake_case ( ): _UpperCamelCase = [] for root, _, files in os.walk(__snake_case ): if "__init__.py" in files: _UpperCamelCase = os.path.join(__snake_case , '''__init__.py''' ) _UpperCamelCase = parse_init(__snake_case ) if objects is not None: _UpperCamelCase = analyze_results(*__snake_case ) if len(__snake_case ) > 0: _UpperCamelCase = f"""Problem in {fname}, both halves do not define the same objects.\n{errors[0]}""" failures.append('''\n'''.join(__snake_case ) ) if len(__snake_case ) > 0: raise ValueError('''\n\n'''.join(__snake_case ) ) def _snake_case ( ): _UpperCamelCase = [] for path, directories, files in os.walk(__snake_case ): for folder in directories: # Ignore private modules if folder.startswith('''_''' ): directories.remove(__snake_case ) continue # Ignore leftovers from branches (empty folders apart from pycache) if len(list((Path(__snake_case ) / folder).glob('''*.py''' ) ) ) == 0: continue _UpperCamelCase = str((Path(__snake_case ) / folder).relative_to(__snake_case ) ) _UpperCamelCase = short_path.replace(os.path.sep , '''.''' ) submodules.append(__snake_case ) for fname in files: if fname == "__init__.py": continue _UpperCamelCase = str((Path(__snake_case ) / fname).relative_to(__snake_case ) ) _UpperCamelCase = short_path.replace('''.py''' , '''''' ).replace(os.path.sep , '''.''' ) if len(submodule.split('''.''' ) ) == 1: submodules.append(__snake_case ) return submodules _lowerCAmelCase = [ "convert_pytorch_checkpoint_to_tf2", "modeling_flax_pytorch_utils", "models.esm.openfold_utils", ] def _snake_case ( ): # This is to make sure the transformers module imported is the one in the repo. from transformers.utils import direct_transformers_import _UpperCamelCase = direct_transformers_import(__snake_case ) _UpperCamelCase = set(transformers._import_structure.keys() ) # This contains all the base keys of the _import_structure object defined in the init, but if the user is missing # some optional dependencies, they may not have all of them. Thus we read the init to read all additions and # (potentiall re-) add them. with open(os.path.join(__snake_case , '''__init__.py''' ) , '''r''' ) as f: _UpperCamelCase = f.read() import_structure_keys.update(set(re.findall(R'''import_structure\[\"([^\"]*)\"\]''' , __snake_case ) ) ) _UpperCamelCase = [ module for module in get_transformers_submodules() if module not in IGNORE_SUBMODULES and module not in import_structure_keys ] if len(__snake_case ) > 0: _UpperCamelCase = '''\n'''.join(f"""- {module}""" for module in module_not_registered ) raise ValueError( '''The following submodules are not properly registed in the main init of Transformers:\n''' f"""{list_of_modules}\n""" '''Make sure they appear somewhere in the keys of `_import_structure` with an empty list as value.''' ) if __name__ == "__main__": check_all_inits() check_submodules()
71
import enum import warnings from .. import MODEL_FOR_CAUSAL_LM_MAPPING, TF_MODEL_FOR_CAUSAL_LM_MAPPING from ..utils import add_end_docstrings, is_tf_available from .base import PIPELINE_INIT_ARGS, Pipeline if is_tf_available(): import tensorflow as tf class lowerCAmelCase_ ( enum.Enum ): UpperCAmelCase = 0 UpperCAmelCase = 1 UpperCAmelCase = 2 @add_end_docstrings(__lowercase ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "\n In 1991, the remains of Russian Tsar Nicholas II and his family (except for Alexei and Maria) are discovered. The\n voice of Nicholas's young son, Tsarevich Alexei Nikolaevich, narrates the remainder of the story. 1883 Western\n Siberia, a young Grigori Rasputin is asked by his father and a group of men to perform magic. Rasputin has a vision\n and denounces one of the men as a horse thief. Although his father initially slaps him for making such an\n accusation, Rasputin watches as the man is chased outside and beaten. Twenty years later, Rasputin sees a vision of\n the Virgin Mary, prompting him to become a priest. Rasputin quickly becomes famous, with people, even a bishop,\n begging for his blessing. <eod> </s> <eos>\n " def __init__( self : Tuple , *_A : List[str] , **_A : str ): super().__init__(*_A , **_A ) self.check_model_type( TF_MODEL_FOR_CAUSAL_LM_MAPPING if self.framework == '''tf''' else MODEL_FOR_CAUSAL_LM_MAPPING ) if "prefix" not in self._preprocess_params: # This is very specific. The logic is quite complex and needs to be done # as a "default". # It also defines both some preprocess_kwargs and generate_kwargs # which is why we cannot put them in their respective methods. _UpperCamelCase = None if self.model.config.prefix is not None: _UpperCamelCase = self.model.config.prefix if prefix is None and self.model.__class__.__name__ in [ "XLNetLMHeadModel", "TransfoXLLMHeadModel", "TFXLNetLMHeadModel", "TFTransfoXLLMHeadModel", ]: # For XLNet and TransformerXL we add an article to the prompt to give more state to the model. _UpperCamelCase = self.XL_PREFIX if prefix is not None: # Recalculate some generate_kwargs linked to prefix. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = self._sanitize_parameters(prefix=_A , **self._forward_params ) _UpperCamelCase = {**self._preprocess_params, **preprocess_params} _UpperCamelCase = {**self._forward_params, **forward_params} def UpperCamelCase_ ( self : Dict , _A : Optional[int]=None , _A : Any=None , _A : Optional[int]=None , _A : List[str]=None , _A : List[Any]=None , _A : int=None , _A : Tuple=None , _A : Optional[Any]=None , **_A : Optional[int] , ): _UpperCamelCase = {} if prefix is not None: _UpperCamelCase = prefix if prefix: _UpperCamelCase = self.tokenizer( _A , padding=_A , add_special_tokens=_A , return_tensors=self.framework ) _UpperCamelCase = prefix_inputs['''input_ids'''].shape[-1] if handle_long_generation is not None: if handle_long_generation not in {"hole"}: raise ValueError( F"""{handle_long_generation} is not a valid value for `handle_long_generation` parameter expected""" ''' [None, \'hole\']''' ) _UpperCamelCase = handle_long_generation preprocess_params.update(_A ) _UpperCamelCase = generate_kwargs _UpperCamelCase = {} if return_full_text is not None and return_type is None: if return_text is not None: raise ValueError('''`return_text` is mutually exclusive with `return_full_text`''' ) if return_tensors is not None: raise ValueError('''`return_full_text` is mutually exclusive with `return_tensors`''' ) _UpperCamelCase = ReturnType.FULL_TEXT if return_full_text else ReturnType.NEW_TEXT if return_tensors is not None and return_type is None: if return_text is not None: raise ValueError('''`return_text` is mutually exclusive with `return_tensors`''' ) _UpperCamelCase = ReturnType.TENSORS if return_type is not None: _UpperCamelCase = return_type if clean_up_tokenization_spaces is not None: _UpperCamelCase = clean_up_tokenization_spaces if stop_sequence is not None: _UpperCamelCase = self.tokenizer.encode(_A , add_special_tokens=_A ) if len(_A ) > 1: warnings.warn( '''Stopping on a multiple token sequence is not yet supported on transformers. The first token of''' ''' the stop sequence will be used as the stop sequence string in the interim.''' ) _UpperCamelCase = stop_sequence_ids[0] return preprocess_params, forward_params, postprocess_params def UpperCamelCase_ ( self : int , *_A : Union[str, Any] , **_A : Union[str, Any] ): # Parse arguments if self.model.__class__.__name__ in ["TransfoXLLMHeadModel"]: kwargs.update({'''add_space_before_punct_symbol''': True} ) return super()._parse_and_tokenize(*_A , **_A ) def __call__( self : List[str] , _A : str , **_A : Any ): return super().__call__(_A , **_A ) def UpperCamelCase_ ( self : Optional[Any] , _A : List[str] , _A : int="" , _A : Optional[Any]=None , **_A : Optional[Any] ): _UpperCamelCase = self.tokenizer( prefix + prompt_text , padding=_A , add_special_tokens=_A , return_tensors=self.framework ) _UpperCamelCase = prompt_text if handle_long_generation == "hole": _UpperCamelCase = inputs['''input_ids'''].shape[-1] if "max_new_tokens" in generate_kwargs: _UpperCamelCase = generate_kwargs['''max_new_tokens'''] else: _UpperCamelCase = generate_kwargs.get('''max_length''' , self.model.config.max_length ) - cur_len if new_tokens < 0: raise ValueError('''We cannot infer how many new tokens are expected''' ) if cur_len + new_tokens > self.tokenizer.model_max_length: _UpperCamelCase = self.tokenizer.model_max_length - new_tokens if keep_length <= 0: raise ValueError( '''We cannot use `hole` to handle this generation the number of desired tokens exceeds the''' ''' models max length''' ) _UpperCamelCase = inputs['''input_ids'''][:, -keep_length:] if "attention_mask" in inputs: _UpperCamelCase = inputs['''attention_mask'''][:, -keep_length:] return inputs def UpperCamelCase_ ( self : Dict , _A : Optional[int] , **_A : str ): _UpperCamelCase = model_inputs['''input_ids'''] _UpperCamelCase = model_inputs.get('''attention_mask''' , _A ) # Allow empty prompts if input_ids.shape[1] == 0: _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = 1 else: _UpperCamelCase = input_ids.shape[0] _UpperCamelCase = model_inputs.pop('''prompt_text''' ) # If there is a prefix, we may need to adjust the generation length. Do so without permanently modifying # generate_kwargs, as some of the parameterization may come from the initialization of the pipeline. _UpperCamelCase = generate_kwargs.pop('''prefix_length''' , 0 ) if prefix_length > 0: _UpperCamelCase = '''max_new_tokens''' in generate_kwargs or ( '''generation_config''' in generate_kwargs and generate_kwargs['''generation_config'''].max_new_tokens is not None ) if not has_max_new_tokens: _UpperCamelCase = generate_kwargs.get('''max_length''' ) or self.model.config.max_length generate_kwargs["max_length"] += prefix_length _UpperCamelCase = '''min_new_tokens''' in generate_kwargs or ( '''generation_config''' in generate_kwargs and generate_kwargs['''generation_config'''].min_new_tokens is not None ) if not has_min_new_tokens and "min_length" in generate_kwargs: generate_kwargs["min_length"] += prefix_length # BS x SL _UpperCamelCase = self.model.generate(input_ids=_A , attention_mask=_A , **_A ) _UpperCamelCase = generated_sequence.shape[0] if self.framework == "pt": _UpperCamelCase = generated_sequence.reshape(_A , out_b // in_b , *generated_sequence.shape[1:] ) elif self.framework == "tf": _UpperCamelCase = tf.reshape(_A , (in_b, out_b // in_b, *generated_sequence.shape[1:]) ) return {"generated_sequence": generated_sequence, "input_ids": input_ids, "prompt_text": prompt_text} def UpperCamelCase_ ( self : List[str] , _A : Dict , _A : Optional[Any]=ReturnType.FULL_TEXT , _A : Dict=True ): _UpperCamelCase = model_outputs['''generated_sequence'''][0] _UpperCamelCase = model_outputs['''input_ids'''] _UpperCamelCase = model_outputs['''prompt_text'''] _UpperCamelCase = generated_sequence.numpy().tolist() _UpperCamelCase = [] for sequence in generated_sequence: if return_type == ReturnType.TENSORS: _UpperCamelCase = {'''generated_token_ids''': sequence} elif return_type in {ReturnType.NEW_TEXT, ReturnType.FULL_TEXT}: # Decode text _UpperCamelCase = self.tokenizer.decode( _A , skip_special_tokens=_A , clean_up_tokenization_spaces=_A , ) # Remove PADDING prompt of the sequence if XLNet or Transfo-XL model is used if input_ids is None: _UpperCamelCase = 0 else: _UpperCamelCase = len( self.tokenizer.decode( input_ids[0] , skip_special_tokens=_A , clean_up_tokenization_spaces=_A , ) ) if return_type == ReturnType.FULL_TEXT: _UpperCamelCase = prompt_text + text[prompt_length:] else: _UpperCamelCase = text[prompt_length:] _UpperCamelCase = {'''generated_text''': all_text} records.append(_A ) return records
71
1
import os import unicodedata from shutil import copyfile from typing import Any, Dict, List, Optional, Tuple import sentencepiece as spm from ...tokenization_utils import AddedToken, PreTrainedTokenizer from ...utils import SPIECE_UNDERLINE, logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = {"vocab_file": "spiece.model"} _lowerCAmelCase = { "vocab_file": { "TsinghuaAI/CPM-Generate": "https://huggingface.co/TsinghuaAI/CPM-Generate/resolve/main/spiece.model", } } class lowerCAmelCase_ ( __lowercase ): def __init__( self : str , _A : Optional[Any] , _A : Tuple=False , _A : List[str]=True , _A : List[str]=False , _A : Union[str, Any]="<s>" , _A : str="</s>" , _A : Any="<unk>" , _A : Union[str, Any]="<sep>" , _A : Dict="<pad>" , _A : int="<cls>" , _A : Dict="<mask>" , _A : Optional[int]=["<eop>", "<eod>"] , _A : Optional[Dict[str, Any]] = None , **_A : str , ): _UpperCamelCase = AddedToken(_A , lstrip=_A , rstrip=_A ) if isinstance(_A , _A ) else mask_token _UpperCamelCase = {} if sp_model_kwargs is None else sp_model_kwargs super().__init__( do_lower_case=_A , remove_space=_A , keep_accents=_A , bos_token=_A , eos_token=_A , unk_token=_A , sep_token=_A , pad_token=_A , cls_token=_A , mask_token=_A , additional_special_tokens=_A , sp_model_kwargs=self.sp_model_kwargs , **_A , ) _UpperCamelCase = 3 _UpperCamelCase = do_lower_case _UpperCamelCase = remove_space _UpperCamelCase = keep_accents _UpperCamelCase = vocab_file _UpperCamelCase = spm.SentencePieceProcessor(**self.sp_model_kwargs ) self.sp_model.Load(_A ) try: import jieba except ModuleNotFoundError as error: raise error.__class__( '''You need to install jieba to use CpmTokenizer or CpmTokenizerFast. ''' '''See https://pypi.org/project/jieba/ for installation.''' ) _UpperCamelCase = jieba _UpperCamelCase = str.maketrans(''' \n''' , '''\u2582\u2583''' ) @property # Copied from transformers.models.xlnet.tokenization_xlnet.XLNetTokenizer.vocab_size def UpperCamelCase_ ( self : Dict ): return len(self.sp_model ) def UpperCamelCase_ ( self : str ): _UpperCamelCase = {self.convert_ids_to_tokens(_A ): i for i in range(self.vocab_size )} vocab.update(self.added_tokens_encoder ) return vocab def __getstate__( self : str ): _UpperCamelCase = self.__dict__.copy() _UpperCamelCase = None return state def __setstate__( self : Any , _A : List[str] ): _UpperCamelCase = d # for backward compatibility if not hasattr(self , '''sp_model_kwargs''' ): _UpperCamelCase = {} _UpperCamelCase = spm.SentencePieceProcessor(**self.sp_model_kwargs ) self.sp_model.Load(self.vocab_file ) def UpperCamelCase_ ( self : Tuple , _A : List[Any] ): if self.remove_space: _UpperCamelCase = ''' '''.join(inputs.strip().split() ) else: _UpperCamelCase = inputs _UpperCamelCase = outputs.replace('''``''' , '''"''' ).replace('''\'\'''' , '''"''' ) if not self.keep_accents: _UpperCamelCase = unicodedata.normalize('''NFKD''' , _A ) _UpperCamelCase = ''''''.join([c for c in outputs if not unicodedata.combining(_A )] ) if self.do_lower_case: _UpperCamelCase = outputs.lower() return outputs def UpperCamelCase_ ( self : Optional[int] , _A : str ): _UpperCamelCase = self.preprocess_text(_A ) _UpperCamelCase = self.sp_model.encode(_A , out_type=_A ) _UpperCamelCase = [] for piece in pieces: if len(_A ) > 1 and piece[-1] == str(''',''' ) and piece[-2].isdigit(): _UpperCamelCase = self.sp_model.EncodeAsPieces(piece[:-1].replace(_A , '''''' ) ) if piece[0] != SPIECE_UNDERLINE and cur_pieces[0][0] == SPIECE_UNDERLINE: if len(cur_pieces[0] ) == 1: _UpperCamelCase = cur_pieces[1:] else: _UpperCamelCase = cur_pieces[0][1:] cur_pieces.append(piece[-1] ) new_pieces.extend(_A ) else: new_pieces.append(_A ) return new_pieces def UpperCamelCase_ ( self : List[Any] , _A : str ): return self.sp_model.PieceToId(_A ) def UpperCamelCase_ ( self : int , _A : Union[str, Any] ): return self.sp_model.IdToPiece(_A ) def UpperCamelCase_ ( self : Optional[Any] , _A : List[Any] ): _UpperCamelCase = ''''''.join(_A ).replace(_A , ''' ''' ).strip() return out_string def UpperCamelCase_ ( self : Optional[int] , _A : List[int] , _A : Optional[List[int]] = None ): _UpperCamelCase = [self.sep_token_id] _UpperCamelCase = [self.cls_token_id] if token_ids_a is None: return token_ids_a + sep + cls return token_ids_a + sep + token_ids_a + sep + cls def UpperCamelCase_ ( self : int , _A : List[int] , _A : Optional[List[int]] = None , _A : bool = False ): if already_has_special_tokens: return super().get_special_tokens_mask( token_ids_a=_A , token_ids_a=_A , already_has_special_tokens=_A ) if token_ids_a is not None: return ([0] * len(_A )) + [1] + ([0] * len(_A )) + [1, 1] return ([0] * len(_A )) + [1, 1] def UpperCamelCase_ ( self : Optional[int] , _A : List[int] , _A : Optional[List[int]] = None ): _UpperCamelCase = [self.sep_token_id] _UpperCamelCase = [2] if token_ids_a is None: return len(token_ids_a + sep ) * [0] + cls_segment_id return len(token_ids_a + sep ) * [0] + len(token_ids_a + sep ) * [1] + cls_segment_id def UpperCamelCase_ ( self : List[Any] , _A : str , _A : Optional[str] = None ): if not os.path.isdir(_A ): logger.error(F"""Vocabulary path ({save_directory}) should be a directory""" ) return _UpperCamelCase = os.path.join( _A , (filename_prefix + '''-''' if filename_prefix else '''''') + VOCAB_FILES_NAMES['''vocab_file'''] ) if os.path.abspath(self.vocab_file ) != os.path.abspath(_A ) and os.path.isfile(self.vocab_file ): copyfile(self.vocab_file , _A ) elif not os.path.isfile(self.vocab_file ): with open(_A , '''wb''' ) as fi: _UpperCamelCase = self.sp_model.serialized_model_proto() fi.write(_A ) return (out_vocab_file,) def UpperCamelCase_ ( self : Dict , *_A : List[str] , **_A : Any ): _UpperCamelCase = super()._decode(*_A , **_A ) _UpperCamelCase = text.replace(''' ''' , '''''' ).replace('''\u2582''' , ''' ''' ).replace('''\u2583''' , '''\n''' ) return text
71
import unittest from transformers import is_torch_available from transformers.testing_utils import require_sentencepiece, require_tokenizers, require_torch, slow, torch_device if is_torch_available(): from transformers import AutoModelForSeqaSeqLM, AutoTokenizer @require_torch @require_sentencepiece @require_tokenizers class lowerCAmelCase_ ( unittest.TestCase ): @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = AutoModelForSeqaSeqLM.from_pretrained('''google/mt5-small''' , return_dict=_A ).to(_A ) _UpperCamelCase = AutoTokenizer.from_pretrained('''google/mt5-small''' ) _UpperCamelCase = tokenizer('''Hello there''' , return_tensors='''pt''' ).input_ids _UpperCamelCase = tokenizer('''Hi I am''' , return_tensors='''pt''' ).input_ids _UpperCamelCase = model(input_ids.to(_A ) , labels=labels.to(_A ) ).loss _UpperCamelCase = -(labels.shape[-1] * loss.item()) _UpperCamelCase = -84.9127 self.assertTrue(abs(mtf_score - EXPECTED_SCORE ) < 1e-4 )
71
1
import json import sys def _snake_case ( __snake_case , __snake_case ): with open(__snake_case , encoding='''utf-8''' ) as f: _UpperCamelCase = json.load(__snake_case ) _UpperCamelCase = ['''<details>''', '''<summary>Show updated benchmarks!</summary>''', ''' '''] for benchmark_name in sorted(__snake_case ): _UpperCamelCase = results[benchmark_name] _UpperCamelCase = benchmark_name.split('''/''' )[-1] output_md.append(f"""### Benchmark: {benchmark_file_name}""" ) _UpperCamelCase = '''| metric |''' _UpperCamelCase = '''|--------|''' _UpperCamelCase = '''| new / old (diff) |''' for metric_name in sorted(__snake_case ): _UpperCamelCase = benchmark_res[metric_name] _UpperCamelCase = metric_vals['''new'''] _UpperCamelCase = metric_vals.get('''old''' , __snake_case ) _UpperCamelCase = metric_vals.get('''diff''' , __snake_case ) _UpperCamelCase = f""" {new_val:f}""" if isinstance(__snake_case , (int, float) ) else '''None''' if old_val is not None: val_str += f""" / {old_val:f}""" if isinstance(__snake_case , (int, float) ) else "None" if dif_val is not None: val_str += f""" ({dif_val:f})""" if isinstance(__snake_case , (int, float) ) else "None" title += " " + metric_name + " |" lines += "---|" value += val_str + " |" output_md += [title, lines, value, " "] output_md.append('''</details>''' ) with open(__snake_case , '''w''' , encoding='''utf-8''' ) as f: f.writelines('''\n'''.join(__snake_case ) ) if __name__ == "__main__": _lowerCAmelCase = sys.argv[1] _lowerCAmelCase = sys.argv[2] format_json_to_md(input_json_file, output_md_file)
71
import logging import os import sys from dataclasses import dataclass, field from typing import Optional from seqaseq_trainer import SeqaSeqTrainer from seqaseq_training_args import SeqaSeqTrainingArguments import transformers from transformers import ( AutoConfig, AutoModelForSeqaSeqLM, AutoTokenizer, HfArgumentParser, MBartTokenizer, MBartTokenizerFast, set_seed, ) from transformers.trainer_utils import EvaluationStrategy, is_main_process from transformers.training_args import ParallelMode from utils import ( SeqaSeqDataCollator, SeqaSeqDataset, assert_all_frozen, build_compute_metrics_fn, check_output_dir, freeze_embeds, freeze_params, lmap, save_json, use_task_specific_params, write_txt_file, ) _lowerCAmelCase = logging.getLogger(__name__) @dataclass class lowerCAmelCase_ : UpperCAmelCase = field( metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Pretrained config name or path if not the same as model_name"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"}, ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Whether tp freeze the encoder."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Whether to freeze the embeddings."} ) @dataclass class lowerCAmelCase_ : UpperCAmelCase = field( metadata={"help": "The input data dir. Should contain the .tsv files (or other data files) for the task."} ) UpperCAmelCase = field( default="summarization", metadata={"help": "Task name, summarization (or summarization_{dataset} for pegasus) or translation"}, ) UpperCAmelCase = field( default=1024, metadata={ "help": ( "The maximum total input sequence length after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field( default=128, metadata={ "help": ( "The maximum total sequence length for target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field( default=142, metadata={ "help": ( "The maximum total sequence length for validation target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded. " "This argument is also used to override the ``max_length`` param of ``model.generate``, which is used " "during ``evaluate`` and ``predict``." ) }, ) UpperCAmelCase = field( default=142, metadata={ "help": ( "The maximum total sequence length for test target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field(default=-1, metadata={"help": "# training examples. -1 means use all."} ) UpperCAmelCase = field(default=-1, metadata={"help": "# validation examples. -1 means use all."} ) UpperCAmelCase = field(default=-1, metadata={"help": "# test examples. -1 means use all."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Source language id for translation."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Target language id for translation."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "# num_beams to use for evaluation."} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "If only pad tokens should be ignored. This assumes that `config.pad_token_id` is defined."}, ) def _snake_case ( __snake_case , __snake_case , __snake_case ): logger.info(f"""***** {split} metrics *****""" ) for key in sorted(metrics.keys() ): logger.info(f""" {key} = {metrics[key]}""" ) save_json(__snake_case , os.path.join(__snake_case , f"""{split}_results.json""" ) ) def _snake_case ( ): # See all possible arguments in src/transformers/training_args.py # or by passing the --help flag to this script. # We now keep distinct sets of args, for a cleaner separation of concerns. _UpperCamelCase = HfArgumentParser((ModelArguments, DataTrainingArguments, SeqaSeqTrainingArguments) ) if len(sys.argv ) == 2 and sys.argv[1].endswith('''.json''' ): # If we pass only one argument to the script and it's the path to a json file, # let's parse it to get our arguments. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1] ) ) else: _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = parser.parse_args_into_dataclasses() check_output_dir(__snake_case ) # Setup logging logging.basicConfig( format='''%(asctime)s - %(levelname)s - %(name)s - %(message)s''' , datefmt='''%m/%d/%Y %H:%M:%S''' , level=logging.INFO if training_args.local_rank in [-1, 0] else logging.WARN , ) logger.warning( '''Process rank: %s, device: %s, n_gpu: %s, distributed training: %s, 16-bits training: %s''' , training_args.local_rank , training_args.device , training_args.n_gpu , bool(training_args.parallel_mode == ParallelMode.DISTRIBUTED ) , training_args.fpaa , ) transformers.utils.logging.enable_default_handler() transformers.utils.logging.enable_explicit_format() # Set the verbosity to info of the Transformers logger (on main process only): if is_main_process(training_args.local_rank ): transformers.utils.logging.set_verbosity_info() logger.info('''Training/evaluation parameters %s''' , __snake_case ) # Set seed set_seed(training_args.seed ) # Load pretrained model and tokenizer # # Distributed training: # The .from_pretrained methods guarantee that only one local process can concurrently # download model & vocab. _UpperCamelCase = AutoConfig.from_pretrained( model_args.config_name if model_args.config_name else model_args.model_name_or_path , cache_dir=model_args.cache_dir , ) _UpperCamelCase = ('''encoder_layerdrop''', '''decoder_layerdrop''', '''dropout''', '''attention_dropout''') for p in extra_model_params: if getattr(__snake_case , __snake_case , __snake_case ): assert hasattr(__snake_case , __snake_case ), f"""({config.__class__.__name__}) doesn't have a `{p}` attribute""" setattr(__snake_case , __snake_case , getattr(__snake_case , __snake_case ) ) _UpperCamelCase = AutoTokenizer.from_pretrained( model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path , cache_dir=model_args.cache_dir , ) _UpperCamelCase = AutoModelForSeqaSeqLM.from_pretrained( model_args.model_name_or_path , from_tf='''.ckpt''' in model_args.model_name_or_path , config=__snake_case , cache_dir=model_args.cache_dir , ) # use task specific params use_task_specific_params(__snake_case , data_args.task ) # set num_beams for evaluation if data_args.eval_beams is None: _UpperCamelCase = model.config.num_beams # set decoder_start_token_id for MBart if model.config.decoder_start_token_id is None and isinstance(__snake_case , (MBartTokenizer, MBartTokenizerFast) ): assert ( data_args.tgt_lang is not None and data_args.src_lang is not None ), "mBart requires --tgt_lang and --src_lang" if isinstance(__snake_case , __snake_case ): _UpperCamelCase = tokenizer.lang_code_to_id[data_args.tgt_lang] else: _UpperCamelCase = tokenizer.convert_tokens_to_ids(data_args.tgt_lang ) if model_args.freeze_embeds: freeze_embeds(__snake_case ) if model_args.freeze_encoder: freeze_params(model.get_encoder() ) assert_all_frozen(model.get_encoder() ) _UpperCamelCase = SeqaSeqDataset # Get datasets _UpperCamelCase = ( dataset_class( __snake_case , type_path='''train''' , data_dir=data_args.data_dir , n_obs=data_args.n_train , max_target_length=data_args.max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_train else None ) _UpperCamelCase = ( dataset_class( __snake_case , type_path='''val''' , data_dir=data_args.data_dir , n_obs=data_args.n_val , max_target_length=data_args.val_max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_eval or training_args.evaluation_strategy != EvaluationStrategy.NO else None ) _UpperCamelCase = ( dataset_class( __snake_case , type_path='''test''' , data_dir=data_args.data_dir , n_obs=data_args.n_test , max_target_length=data_args.test_max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_predict else None ) # Initialize our Trainer _UpperCamelCase = ( build_compute_metrics_fn(data_args.task , __snake_case ) if training_args.predict_with_generate else None ) _UpperCamelCase = SeqaSeqTrainer( model=__snake_case , args=__snake_case , data_args=__snake_case , train_dataset=__snake_case , eval_dataset=__snake_case , data_collator=SeqaSeqDataCollator( __snake_case , __snake_case , model.config.decoder_start_token_id , training_args.tpu_num_cores ) , compute_metrics=__snake_case , tokenizer=__snake_case , ) _UpperCamelCase = {} # Training if training_args.do_train: logger.info('''*** Train ***''' ) _UpperCamelCase = trainer.train( model_path=model_args.model_name_or_path if os.path.isdir(model_args.model_name_or_path ) else None ) _UpperCamelCase = train_result.metrics _UpperCamelCase = data_args.n_train trainer.save_model() # this also saves the tokenizer if trainer.is_world_process_zero(): handle_metrics('''train''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) # Need to save the state, since Trainer.save_model saves only the tokenizer with the model trainer.state.save_to_json(os.path.join(training_args.output_dir , '''trainer_state.json''' ) ) # For convenience, we also re-save the tokenizer to the same directory, # so that you can share your model easily on huggingface.co/models =) tokenizer.save_pretrained(training_args.output_dir ) # Evaluation if training_args.do_eval: logger.info('''*** Evaluate ***''' ) _UpperCamelCase = trainer.evaluate(metric_key_prefix='''val''' ) _UpperCamelCase = data_args.n_val _UpperCamelCase = round(metrics['''val_loss'''] , 4 ) if trainer.is_world_process_zero(): handle_metrics('''val''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) if training_args.do_predict: logger.info('''*** Predict ***''' ) _UpperCamelCase = trainer.predict(test_dataset=__snake_case , metric_key_prefix='''test''' ) _UpperCamelCase = test_output.metrics _UpperCamelCase = data_args.n_test if trainer.is_world_process_zero(): _UpperCamelCase = round(metrics['''test_loss'''] , 4 ) handle_metrics('''test''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) if training_args.predict_with_generate: _UpperCamelCase = tokenizer.batch_decode( test_output.predictions , skip_special_tokens=__snake_case , clean_up_tokenization_spaces=__snake_case ) _UpperCamelCase = lmap(str.strip , __snake_case ) write_txt_file(__snake_case , os.path.join(training_args.output_dir , '''test_generations.txt''' ) ) if trainer.is_world_process_zero(): save_json(__snake_case , os.path.join(training_args.output_dir , '''all_results.json''' ) ) return all_metrics def _snake_case ( __snake_case ): # For xla_spawn (TPUs) main() if __name__ == "__main__": main()
71
1
import argparse import json import os import re import shutil import torch from transformers import BioGptConfig, BioGptForCausalLM from transformers.models.biogpt.tokenization_biogpt import VOCAB_FILES_NAMES from transformers.tokenization_utils_base import TOKENIZER_CONFIG_FILE from transformers.utils import WEIGHTS_NAME, logging logging.set_verbosity_warning() _lowerCAmelCase = 2 class lowerCAmelCase_ : def __init__( self : Optional[Any] , *, # begin keyword-only arguments _A : Optional[Any]="<s>" , _A : Union[str, Any]="<pad>" , _A : Dict="</s>" , _A : Optional[Any]="<unk>" , _A : int=None , ): _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = bos, unk, pad, eos _UpperCamelCase = [] _UpperCamelCase = [] _UpperCamelCase = {} _UpperCamelCase = self.add_symbol(_A ) _UpperCamelCase = self.add_symbol(_A ) _UpperCamelCase = self.add_symbol(_A ) _UpperCamelCase = self.add_symbol(_A ) if extra_special_symbols: for s in extra_special_symbols: self.add_symbol(_A ) _UpperCamelCase = len(self.symbols ) def __eq__( self : Optional[Any] , _A : int ): return self.indices == other.indices def __getitem__( self : Tuple , _A : Tuple ): if idx < len(self.symbols ): return self.symbols[idx] return self.unk_word def __len__( self : List[Any] ): return len(self.symbols ) def __contains__( self : Union[str, Any] , _A : Optional[Any] ): return sym in self.indices @classmethod def UpperCamelCase_ ( cls : List[str] , _A : Any ): _UpperCamelCase = cls() d.add_from_file(_A ) return d def UpperCamelCase_ ( self : List[Any] , _A : int , _A : Union[str, Any]=1 , _A : Union[str, Any]=False ): if word in self.indices and not overwrite: _UpperCamelCase = self.indices[word] _UpperCamelCase = self.count[idx] + n return idx else: _UpperCamelCase = len(self.symbols ) _UpperCamelCase = idx self.symbols.append(_A ) self.count.append(_A ) return idx def UpperCamelCase_ ( self : Union[str, Any] , _A : Optional[int] ): return 0 def UpperCamelCase_ ( self : Tuple , _A : Dict ): if isinstance(_A , _A ): try: with open(_A , '''r''' , encoding='''utf-8''' ) as fd: self.add_from_file(_A ) except FileNotFoundError as fnfe: raise fnfe except UnicodeError: raise Exception('''Incorrect encoding detected in {}, please rebuild the dataset'''.format(_A ) ) return _UpperCamelCase = f.readlines() _UpperCamelCase = self._load_meta(_A ) for line in lines[indices_start_line:]: try: _UpperCamelCase , _UpperCamelCase = line.rstrip().rsplit(''' ''' , 1 ) if field == "#fairseq:overwrite": _UpperCamelCase = True _UpperCamelCase , _UpperCamelCase = line.rsplit(''' ''' , 1 ) else: _UpperCamelCase = False _UpperCamelCase = int(_A ) _UpperCamelCase = line if word in self and not overwrite: raise RuntimeError( '''Duplicate word found when loading Dictionary: \'{}\'. ''' '''Duplicate words can overwrite earlier ones by adding the ''' '''#fairseq:overwrite flag at the end of the corresponding row ''' '''in the dictionary file. If using the Camembert model, please ''' '''download an updated copy of the model file.'''.format(_A ) ) self.add_symbol(_A , n=_A , overwrite=_A ) except ValueError: raise ValueError('''Incorrect dictionary format, expected \'<token> <cnt> [flags]\'''' ) def _snake_case ( __snake_case ): # (1) remove word breaking symbol, (2) add word ending symbol where the word is not broken up, # e.g.: d = {'le@@': 5, 'tt@@': 6, 'er': 7} => {'le': 5, 'tt': 6, 'er</w>': 7} _UpperCamelCase = dict((re.sub(R'''@@$''' , '''''' , __snake_case ), v) if k.endswith('''@@''' ) else (re.sub(R'''$''' , '''</w>''' , __snake_case ), v) for k, v in d.items() ) _UpperCamelCase = '''<s> <pad> </s> <unk>'''.split() # restore the special tokens for k in keep_keys: del da[f"""{k}</w>"""] _UpperCamelCase = d[k] # restore return da def _snake_case ( __snake_case , __snake_case ): # prep if not os.path.exists(__snake_case ): raise ValueError(f"""path {biogpt_checkpoint_path} does not exist!""" ) os.makedirs(__snake_case , exist_ok=__snake_case ) print(f"""Writing results to {pytorch_dump_folder_path}""" ) # handle various types of models _UpperCamelCase = os.path.join(__snake_case , '''checkpoint.pt''' ) if not os.path.isfile(__snake_case ): raise ValueError(f"""path to the file {checkpoint_file} does not exist!""" ) _UpperCamelCase = torch.load(__snake_case , map_location='''cpu''' ) _UpperCamelCase = chkpt['''cfg''']['''model'''] # dicts _UpperCamelCase = os.path.join(__snake_case , '''dict.txt''' ) if not os.path.isfile(__snake_case ): raise ValueError(f"""path to the file {dict_file} does not exist!""" ) _UpperCamelCase = Dictionary.load(__snake_case ) _UpperCamelCase = rewrite_dict_keys(src_dict.indices ) _UpperCamelCase = len(__snake_case ) _UpperCamelCase = os.path.join(__snake_case , VOCAB_FILES_NAMES['''vocab_file'''] ) print(f"""Generating {src_vocab_file} of {src_vocab_size} records""" ) with open(__snake_case , '''w''' , encoding='''utf-8''' ) as f: f.write(json.dumps(__snake_case , ensure_ascii=__snake_case , indent=__snake_case ) ) # merges_file (bpecodes) _UpperCamelCase = os.path.join(__snake_case , '''bpecodes''' ) if not os.path.isfile(__snake_case ): raise ValueError(f"""path to the file {bpecodes_file} does not exist!""" ) _UpperCamelCase = os.path.join(__snake_case , VOCAB_FILES_NAMES['''merges_file'''] ) shutil.copyfile(__snake_case , __snake_case ) # model config _UpperCamelCase = os.path.join(__snake_case , '''config.json''' ) _UpperCamelCase = { '''activation_dropout''': args['''activation_dropout'''], '''architectures''': ['''BioGptForCausalLM'''], '''attention_probs_dropout_prob''': args['''attention_dropout'''], '''bos_token_id''': 0, '''eos_token_id''': 2, '''hidden_act''': args['''activation_fn'''], '''hidden_dropout_prob''': args['''dropout'''], '''hidden_size''': args['''decoder_embed_dim'''], '''initializer_range''': 0.02, '''intermediate_size''': args['''decoder_ffn_embed_dim'''], '''layer_norm_eps''': 1E-12, '''layerdrop''': args['''decoder_layerdrop'''], '''max_position_embeddings''': args['''max_target_positions'''], '''model_type''': '''biogpt''', '''num_attention_heads''': args['''decoder_attention_heads'''], '''num_hidden_layers''': args['''decoder_layers'''], '''pad_token_id''': 1, '''scale_embedding''': not args['''no_scale_embedding'''], '''tie_word_embeddings''': args['''share_decoder_input_output_embed'''], '''vocab_size''': src_vocab_size, } # good hparam defaults to start with print(f"""Generating {biogpt_model_config_file}""" ) with open(__snake_case , '''w''' , encoding='''utf-8''' ) as f: f.write(json.dumps(__snake_case , ensure_ascii=__snake_case , indent=__snake_case ) ) # tokenizer config _UpperCamelCase = os.path.join(__snake_case , __snake_case ) _UpperCamelCase = { '''bos_token''': '''<s>''', '''eos_token''': '''</s>''', '''model_max_length''': 1024, '''pad_token''': '''<pad>''', '''special_tokens_map_file''': None, '''tokenizer_class''': '''BioGptTokenizer''', '''unk_token''': '''<unk>''', } print(f"""Generating {biogpt_tokenizer_config_file}""" ) with open(__snake_case , '''w''' , encoding='''utf-8''' ) as f: f.write(json.dumps(__snake_case , ensure_ascii=__snake_case , indent=__snake_case ) ) # model _UpperCamelCase = chkpt['''model'''] # remove unneeded keys _UpperCamelCase = [ '''decoder.version''', ] for k in ignore_keys: model_state_dict.pop(__snake_case , __snake_case ) _UpperCamelCase = list(model_state_dict.keys() ) for layer_name in layer_names: if layer_name.endswith('''output_projection.weight''' ): _UpperCamelCase = model_state_dict.pop(__snake_case ) else: _UpperCamelCase = model_state_dict.pop(__snake_case ) _UpperCamelCase = BioGptConfig.from_pretrained(__snake_case ) _UpperCamelCase = BioGptForCausalLM(__snake_case ) # check that it loads ok model_new.load_state_dict(__snake_case ) # save _UpperCamelCase = os.path.join(__snake_case , __snake_case ) print(f"""Generating {pytorch_weights_dump_path}""" ) torch.save(__snake_case , __snake_case ) print('''Conversion is done!''' ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() # Required parameters parser.add_argument( "--biogpt_checkpoint_path", default=None, type=str, required=True, help=( "Path to the official PyTorch checkpoint file which is expected to reside in the dump dir with dicts," " bpecodes, etc." ), ) parser.add_argument( "--pytorch_dump_folder_path", default=None, type=str, required=True, help="Path to the output PyTorch model." ) _lowerCAmelCase = parser.parse_args() convert_biogpt_checkpoint_to_pytorch(args.biogpt_checkpoint_path, args.pytorch_dump_folder_path)
71
from __future__ import annotations import typing from collections import Counter def _snake_case ( __snake_case ): _UpperCamelCase = Counter() for base in range(1 , max_perimeter + 1 ): for perpendicular in range(__snake_case , max_perimeter + 1 ): _UpperCamelCase = (base * base + perpendicular * perpendicular) ** 0.5 if hypotenuse == int(__snake_case ): _UpperCamelCase = int(base + perpendicular + hypotenuse ) if perimeter > max_perimeter: continue triplets[perimeter] += 1 return triplets def _snake_case ( __snake_case = 1000 ): _UpperCamelCase = pythagorean_triple(__snake_case ) return triplets.most_common(1 )[0][0] if __name__ == "__main__": print(f'Perimeter {solution()} has maximum solutions')
71
1
import argparse import json import os import torch from torch import nn from transformers import NllbMoeConfig, NllbMoeModel from transformers.modeling_utils import dtype_byte_size from transformers.utils import WEIGHTS_INDEX_NAME, WEIGHTS_NAME def _snake_case ( __snake_case ): _UpperCamelCase = [ '''encoder.version''', '''decoder.version''', '''model.encoder.version''', '''model.decoder.version''', '''decoder.output_projection.weight''', '''_float_tensor''', '''encoder.embed_positions._float_tensor''', '''decoder.embed_positions._float_tensor''', ] for k in ignore_keys: state_dict.pop(__snake_case , __snake_case ) def _snake_case ( __snake_case ): _UpperCamelCase , _UpperCamelCase = emb.weight.shape _UpperCamelCase = nn.Linear(__snake_case , __snake_case , bias=__snake_case ) _UpperCamelCase = emb.weight.data return lin_layer def _snake_case ( __snake_case , __snake_case=None ): _UpperCamelCase = {} for old_key in state_dict.keys(): _UpperCamelCase = old_key if "moe_layer.experts." in key: if expert_idx is not None: _UpperCamelCase = key.replace('''moe_layer.experts.0''' , f"""ffn.experts.expert_{expert_idx}""" ) else: _UpperCamelCase = key.replace('''moe_layer.experts.''' , '''ffn.experts.expert_''' ) if "gate" in key: _UpperCamelCase = key.replace('''.moe_layer.gate.wg''' , '''.ffn.router.classifier''' ) if "fc2" and "experts" not in key: _UpperCamelCase = key.replace('''.fc2.''' , '''.ffn.fc2.''' ) if "fc1" and "experts" not in key: _UpperCamelCase = key.replace('''.fc1.''' , '''.ffn.fc1.''' ) if ".encoder_attn." in key: _UpperCamelCase = key.replace('''.encoder_attn.''' , '''.cross_attention.''' ) if "encoder_attn_layer_norm" in key: _UpperCamelCase = key.replace('''encoder_attn_layer_norm''' , '''cross_attention_layer_norm''' ) if "final_layer_norm" in key: _UpperCamelCase = key.replace('''final_layer_norm''' , '''ff_layer_norm''' ) _UpperCamelCase = state_dict[old_key] return new_dict def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case = WEIGHTS_NAME ): _UpperCamelCase = [] _UpperCamelCase = 0 os.makedirs(__snake_case , exist_ok=__snake_case ) for expert in range(__snake_case ): _UpperCamelCase = switch_checkpoint_path + f"""-rank-{expert}.pt""" if os.path.isfile(__snake_case ): _UpperCamelCase = torch.load(__snake_case )['''model'''] remove_ignore_keys_(__snake_case ) _UpperCamelCase = rename_fairseq_keys(__snake_case , __snake_case ) _UpperCamelCase = os.path.join( __snake_case , weights_name.replace('''.bin''' , f"""-{len(__snake_case )+1:05d}-of-???.bin""" ) ) torch.save(__snake_case , __snake_case ) sharded_state_dicts.append(expert_state.keys() ) total_size += sum([value.numel() for key, value in expert_state.items()] ) * dtype_byte_size( expert_state[list(__snake_case )[0]].dtype ) # Add the last block _UpperCamelCase = os.path.join(__snake_case , weights_name.replace('''.bin''' , f"""-{len(__snake_case )+1:05d}-of-???.bin""" ) ) _UpperCamelCase = torch.load(switch_checkpoint_path + '''-shared.pt''' )['''model'''] remove_ignore_keys_(__snake_case ) _UpperCamelCase = rename_fairseq_keys(__snake_case , __snake_case ) _UpperCamelCase = shared_weights['''decoder.embed_tokens.weight'''] sharded_state_dicts.append(shared_weights.keys() ) # If we only have the shared weights (dummy model/experts saved on the same file) if len(__snake_case ) == 1: _UpperCamelCase = os.path.join(__snake_case , __snake_case ) torch.save(__snake_case , __snake_case ) return {weights_name: sharded_state_dicts[0]}, None else: torch.save(__snake_case , __snake_case ) # Otherwise, let's build the index _UpperCamelCase = {} for idx, shard in enumerate(__snake_case ): _UpperCamelCase = weights_name.replace('''.bin''' , f"""-{idx+1:05d}-of-{len(__snake_case ):05d}.bin""" ) _UpperCamelCase = os.path.join(__snake_case , weights_name.replace('''.bin''' , f"""-{idx+1:05d}-of-???.bin""" ) ) os.rename(__snake_case , os.path.join(__snake_case , __snake_case ) ) for key in shard: _UpperCamelCase = shard_file # Add the metadata _UpperCamelCase = {'''total_size''': total_size} _UpperCamelCase = {'''metadata''': metadata, '''weight_map''': weight_map} with open(os.path.join(__snake_case , __snake_case ) , '''w''' , encoding='''utf-8''' ) as f: _UpperCamelCase = json.dumps(__snake_case , indent=2 , sort_keys=__snake_case ) + '''\n''' f.write(__snake_case ) return metadata, index if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() # Required parameters parser.add_argument( "--nllb_moe_checkpoint_path", default="/home/arthur_huggingface_co/fairseq/weights/checkpoints/model_moe_54b/checkpoint_2_300000", type=str, required=False, help="Path to a directory containing a folder per layer. Follows the original Google format.", ) parser.add_argument("--dtype", default="float32", type=str, required=False, help="dtype of the saved model") parser.add_argument( "--pytorch_dump_folder_path", default="/home/arthur_huggingface_co/fairseq/weights/checkpoints/hf-converted-moe-54b", type=str, required=False, help="Path to the output pytorch model.", ) _lowerCAmelCase = parser.parse_args() _lowerCAmelCase, _lowerCAmelCase = shard_on_the_fly( args.nllb_moe_checkpoint_path, args.pytorch_dump_folder_path, 128, args.dtype, ) _lowerCAmelCase = NllbMoeConfig.from_pretrained( "facebook/nllb-200-3.3B", encoder_sparse_step=4, decoder_sparse_step=4, num_experts=128 ) config.save_pretrained(args.pytorch_dump_folder_path) _lowerCAmelCase = NllbMoeModel.from_pretrained(args.pytorch_dump_folder_path) print("Done") model.save_pretrained(args.pytorch_dump_folder_path)
71
import torch from diffusers import DPMSolverSDEScheduler from diffusers.utils import torch_device from diffusers.utils.testing_utils import require_torchsde from .test_schedulers import SchedulerCommonTest @require_torchsde class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = (DPMSolverSDEScheduler,) UpperCAmelCase = 10 def UpperCamelCase_ ( self : Tuple , **_A : Union[str, Any] ): _UpperCamelCase = { '''num_train_timesteps''': 1100, '''beta_start''': 0.0001, '''beta_end''': 0.02, '''beta_schedule''': '''linear''', '''noise_sampler_seed''': 0, } config.update(**_A ) return config def UpperCamelCase_ ( self : List[Any] ): for timesteps in [10, 50, 100, 1000]: self.check_over_configs(num_train_timesteps=_A ) def UpperCamelCase_ ( self : List[Any] ): for beta_start, beta_end in zip([0.0_0001, 0.0001, 0.001] , [0.0002, 0.002, 0.02] ): self.check_over_configs(beta_start=_A , beta_end=_A ) def UpperCamelCase_ ( self : List[str] ): for schedule in ["linear", "scaled_linear"]: self.check_over_configs(beta_schedule=_A ) def UpperCamelCase_ ( self : Union[str, Any] ): for prediction_type in ["epsilon", "v_prediction"]: self.check_over_configs(prediction_type=_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for i, t in enumerate(scheduler.timesteps ): _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 167.47_8210_4492_1875 ) < 1e-2 assert abs(result_mean.item() - 0.2178_7059_6456_5277 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 171.59_3521_1181_6406 ) < 1e-2 assert abs(result_mean.item() - 0.2_2342_9068_9229_9652 ) < 1e-3 else: assert abs(result_sum.item() - 162.52_3834_2285_1562 ) < 1e-2 assert abs(result_mean.item() - 0.211_6195_7085_1326 ) < 1e-3 def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config(prediction_type='''v_prediction''' ) _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for i, t in enumerate(scheduler.timesteps ): _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 124.77_1492_0043_9453 ) < 1e-2 assert abs(result_mean.item() - 0.1_6226_2890_1481_6284 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 128.1_6633_6059_5703 ) < 1e-2 assert abs(result_mean.item() - 0.1_6688_3260_0116_7297 ) < 1e-3 else: assert abs(result_sum.item() - 119.8_4875_4882_8125 ) < 1e-2 assert abs(result_mean.item() - 0.1560_5306_6253_6621 ) < 1e-3 def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps , device=_A ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter.to(_A ) * scheduler.init_noise_sigma for t in scheduler.timesteps: _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 167.46_9573_9746_0938 ) < 1e-2 assert abs(result_mean.item() - 0.2_1805_9346_0798_2635 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 171.59_3536_3769_5312 ) < 1e-2 assert abs(result_mean.item() - 0.2_2342_9083_8241_5771 ) < 1e-3 else: assert abs(result_sum.item() - 162.52_3834_2285_1562 ) < 1e-2 assert abs(result_mean.item() - 0.211_6195_7085_1326 ) < 1e-3 def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A , use_karras_sigmas=_A ) scheduler.set_timesteps(self.num_inference_steps , device=_A ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter.to(_A ) * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for t in scheduler.timesteps: _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 176.66_9741_3574_2188 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 177.63_6535_6445_3125 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2 else: assert abs(result_sum.item() - 170.3_1352_2338_8672 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2
71
1
import argparse import os from io import BytesIO from pathlib import Path import requests from clip_retrieval.clip_client import ClipClient from PIL import Image from tqdm import tqdm def _snake_case ( __snake_case , __snake_case , __snake_case ): _UpperCamelCase = 1.5 _UpperCamelCase = int(factor * num_class_images ) _UpperCamelCase = ClipClient( url='''https://knn.laion.ai/knn-service''' , indice_name='''laion_400m''' , num_images=__snake_case , aesthetic_weight=0.1 ) os.makedirs(f"""{class_data_dir}/images""" , exist_ok=__snake_case ) if len(list(Path(f"""{class_data_dir}/images""" ).iterdir() ) ) >= num_class_images: return while True: _UpperCamelCase = client.query(text=__snake_case ) if len(__snake_case ) >= factor * num_class_images or num_images > 1E4: break else: _UpperCamelCase = int(factor * num_images ) _UpperCamelCase = ClipClient( url='''https://knn.laion.ai/knn-service''' , indice_name='''laion_400m''' , num_images=__snake_case , aesthetic_weight=0.1 , ) _UpperCamelCase = 0 _UpperCamelCase = 0 _UpperCamelCase = tqdm(desc='''downloading real regularization images''' , total=__snake_case ) with open(f"""{class_data_dir}/caption.txt""" , '''w''' ) as fa, open(f"""{class_data_dir}/urls.txt""" , '''w''' ) as fa, open( f"""{class_data_dir}/images.txt""" , '''w''' ) as fa: while total < num_class_images: _UpperCamelCase = class_images[count] count += 1 try: _UpperCamelCase = requests.get(images['''url'''] ) if img.status_code == 200: _UpperCamelCase = Image.open(BytesIO(img.content ) ) with open(f"""{class_data_dir}/images/{total}.jpg""" , '''wb''' ) as f: f.write(img.content ) fa.write(images['''caption'''] + '''\n''' ) fa.write(images['''url'''] + '''\n''' ) fa.write(f"""{class_data_dir}/images/{total}.jpg""" + '''\n''' ) total += 1 pbar.update(1 ) else: continue except Exception: continue return def _snake_case ( ): _UpperCamelCase = argparse.ArgumentParser('''''' , add_help=__snake_case ) parser.add_argument('''--class_prompt''' , help='''text prompt to retrieve images''' , required=__snake_case , type=__snake_case ) parser.add_argument('''--class_data_dir''' , help='''path to save images''' , required=__snake_case , type=__snake_case ) parser.add_argument('''--num_class_images''' , help='''number of images to download''' , default=200 , type=__snake_case ) return parser.parse_args() if __name__ == "__main__": _lowerCAmelCase = parse_args() retrieve(args.class_prompt, args.class_data_dir, args.num_class_images)
71
import unittest from typing import Tuple import torch from diffusers.utils import floats_tensor, randn_tensor, torch_all_close, torch_device from diffusers.utils.testing_utils import require_torch @require_torch class lowerCAmelCase_ : @property def UpperCamelCase_ ( self : Optional[int] ): return self.get_dummy_input() @property def UpperCamelCase_ ( self : Dict ): if self.block_type == "down": return (4, 32, 16, 16) elif self.block_type == "mid": return (4, 32, 32, 32) elif self.block_type == "up": return (4, 32, 64, 64) raise ValueError(F"""'{self.block_type}' is not a supported block_type. Set it to 'up', 'mid', or 'down'.""" ) def UpperCamelCase_ ( self : Union[str, Any] , _A : List[str]=True , _A : Any=False , _A : Union[str, Any]=False , _A : int=False , ): _UpperCamelCase = 4 _UpperCamelCase = 32 _UpperCamelCase = (32, 32) _UpperCamelCase = torch.manual_seed(0 ) _UpperCamelCase = torch.device(_A ) _UpperCamelCase = (batch_size, num_channels) + sizes _UpperCamelCase = randn_tensor(_A , generator=_A , device=_A ) _UpperCamelCase = {'''hidden_states''': hidden_states} if include_temb: _UpperCamelCase = 128 _UpperCamelCase = randn_tensor((batch_size, temb_channels) , generator=_A , device=_A ) if include_res_hidden_states_tuple: _UpperCamelCase = torch.manual_seed(1 ) _UpperCamelCase = (randn_tensor(_A , generator=_A , device=_A ),) if include_encoder_hidden_states: _UpperCamelCase = floats_tensor((batch_size, 32, 32) ).to(_A ) if include_skip_sample: _UpperCamelCase = randn_tensor(((batch_size, 3) + sizes) , generator=_A , device=_A ) return dummy_input def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = { '''in_channels''': 32, '''out_channels''': 32, '''temb_channels''': 128, } if self.block_type == "up": _UpperCamelCase = 32 if self.block_type == "mid": init_dict.pop('''out_channels''' ) _UpperCamelCase = self.dummy_input return init_dict, inputs_dict def UpperCamelCase_ ( self : Tuple , _A : Union[str, Any] ): _UpperCamelCase , _UpperCamelCase = self.prepare_init_args_and_inputs_for_common() _UpperCamelCase = self.block_class(**_A ) unet_block.to(_A ) unet_block.eval() with torch.no_grad(): _UpperCamelCase = unet_block(**_A ) if isinstance(_A , _A ): _UpperCamelCase = output[0] self.assertEqual(output.shape , self.output_shape ) _UpperCamelCase = output[0, -1, -3:, -3:] _UpperCamelCase = torch.tensor(_A ).to(_A ) assert torch_all_close(output_slice.flatten() , _A , atol=5e-3 ) @unittest.skipIf(torch_device == '''mps''' , '''Training is not supported in mps''' ) def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase , _UpperCamelCase = self.prepare_init_args_and_inputs_for_common() _UpperCamelCase = self.block_class(**_A ) model.to(_A ) model.train() _UpperCamelCase = model(**_A ) if isinstance(_A , _A ): _UpperCamelCase = output[0] _UpperCamelCase = torch.device(_A ) _UpperCamelCase = randn_tensor(output.shape , device=_A ) _UpperCamelCase = torch.nn.functional.mse_loss(_A , _A ) loss.backward()
71
1
from __future__ import annotations from random import choice def _snake_case ( __snake_case ): return choice(__snake_case ) def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = random_pivot(__snake_case ) # partition based on pivot # linear time _UpperCamelCase = [e for e in lst if e < pivot] _UpperCamelCase = [e for e in lst if e > pivot] # if we get lucky, pivot might be the element we want. # we can easily see this: # small (elements smaller than k) # + pivot (kth element) # + big (elements larger than k) if len(__snake_case ) == k - 1: return pivot # pivot is in elements bigger than k elif len(__snake_case ) < k - 1: return kth_number(__snake_case , k - len(__snake_case ) - 1 ) # pivot is in elements smaller than k else: return kth_number(__snake_case , __snake_case ) if __name__ == "__main__": import doctest doctest.testmod()
71
def _snake_case ( __snake_case ): if not isinstance(__snake_case , __snake_case ): raise TypeError('''Input value must be an \'int\' type''' ) _UpperCamelCase = 0 while number: position += 1 number >>= 1 return position if __name__ == "__main__": import doctest doctest.testmod()
71
1
import unittest from huggingface_hub import hf_hub_download from transformers import MODEL_FOR_VIDEO_CLASSIFICATION_MAPPING, VideoMAEFeatureExtractor from transformers.pipelines import VideoClassificationPipeline, pipeline from transformers.testing_utils import ( is_pipeline_test, nested_simplify, require_decord, require_tf, require_torch, require_torch_or_tf, require_vision, ) from .test_pipelines_common import ANY @is_pipeline_test @require_torch_or_tf @require_vision @require_decord class lowerCAmelCase_ ( unittest.TestCase ): UpperCAmelCase = MODEL_FOR_VIDEO_CLASSIFICATION_MAPPING def UpperCamelCase_ ( self : Optional[Any] , _A : List[str] , _A : Optional[int] , _A : Optional[int] ): _UpperCamelCase = hf_hub_download( repo_id='''nateraw/video-demo''' , filename='''archery.mp4''' , repo_type='''dataset''' ) _UpperCamelCase = VideoClassificationPipeline(model=_A , image_processor=_A , top_k=2 ) _UpperCamelCase = [ example_video_filepath, '''https://huggingface.co/datasets/nateraw/video-demo/resolve/main/archery.mp4''', ] return video_classifier, examples def UpperCamelCase_ ( self : str , _A : Optional[Any] , _A : Optional[int] ): for example in examples: _UpperCamelCase = video_classifier(_A ) self.assertEqual( _A , [ {'''score''': ANY(_A ), '''label''': ANY(_A )}, {'''score''': ANY(_A ), '''label''': ANY(_A )}, ] , ) @require_torch def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = '''hf-internal-testing/tiny-random-VideoMAEForVideoClassification''' _UpperCamelCase = VideoMAEFeatureExtractor( size={'''shortest_edge''': 10} , crop_size={'''height''': 10, '''width''': 10} ) _UpperCamelCase = pipeline( '''video-classification''' , model=_A , feature_extractor=_A , frame_sampling_rate=4 ) _UpperCamelCase = hf_hub_download(repo_id='''nateraw/video-demo''' , filename='''archery.mp4''' , repo_type='''dataset''' ) _UpperCamelCase = video_classifier(_A , top_k=2 ) self.assertEqual( nested_simplify(_A , decimals=4 ) , [{'''score''': 0.5199, '''label''': '''LABEL_0'''}, {'''score''': 0.4801, '''label''': '''LABEL_1'''}] , ) _UpperCamelCase = video_classifier( [ video_file_path, video_file_path, ] , top_k=2 , ) self.assertEqual( nested_simplify(_A , decimals=4 ) , [ [{'''score''': 0.5199, '''label''': '''LABEL_0'''}, {'''score''': 0.4801, '''label''': '''LABEL_1'''}], [{'''score''': 0.5199, '''label''': '''LABEL_0'''}, {'''score''': 0.4801, '''label''': '''LABEL_1'''}], ] , ) @require_tf def UpperCamelCase_ ( self : List[Any] ): pass
71
import argparse import json import os import fairseq import torch from torch import nn from transformers import ( SpeechaTextaConfig, SpeechaTextaForCausalLM, SpeechaTextaTokenizer, SpeechEncoderDecoderConfig, SpeechEncoderDecoderModel, WavaVecaConfig, WavaVecaFeatureExtractor, WavaVecaModel, logging, ) logging.set_verbosity_info() _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "post_extract_proj": "feature_projection.projection", "encoder.pos_conv.0": "encoder.pos_conv_embed.conv", "self_attn.k_proj": "encoder.layers.*.attention.k_proj", "self_attn.v_proj": "encoder.layers.*.attention.v_proj", "self_attn.q_proj": "encoder.layers.*.attention.q_proj", "self_attn.out_proj": "encoder.layers.*.attention.out_proj", "self_attn_layer_norm": "encoder.layers.*.layer_norm", "fc1": "encoder.layers.*.feed_forward.intermediate_dense", "fc2": "encoder.layers.*.feed_forward.output_dense", "final_layer_norm": "encoder.layers.*.final_layer_norm", "encoder.layer_norm": "encoder.layer_norm", "w2v_model.layer_norm": "feature_projection.layer_norm", "quantizer.weight_proj": "quantizer.weight_proj", "quantizer.vars": "quantizer.codevectors", "project_q": "project_q", "final_proj": "project_hid", "w2v_encoder.proj": "lm_head", "mask_emb": "masked_spec_embed", } _lowerCAmelCase = [ "lm_head", "quantizer.weight_proj", "quantizer.codevectors", "project_q", "project_hid", ] def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): for attribute in key.split('''.''' ): _UpperCamelCase = getattr(__snake_case , __snake_case ) if weight_type is not None: _UpperCamelCase = getattr(__snake_case , __snake_case ).shape else: _UpperCamelCase = hf_pointer.shape assert hf_shape == value.shape, ( f"""Shape of hf {key + "." + weight_type if weight_type is not None else ""} is {hf_shape}, but should be""" f""" {value.shape} for {full_name}""" ) if weight_type == "weight": _UpperCamelCase = value elif weight_type == "weight_g": _UpperCamelCase = value elif weight_type == "weight_v": _UpperCamelCase = value elif weight_type == "bias": _UpperCamelCase = value else: _UpperCamelCase = value logger.info(f"""{key + "." + weight_type if weight_type is not None else ""} was initialized from {full_name}.""" ) def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = [] _UpperCamelCase = fairseq_model.state_dict() _UpperCamelCase = hf_model.feature_extractor # if encoder has different dim to decoder -> use proj_weight _UpperCamelCase = None for name, value in fairseq_dict.items(): _UpperCamelCase = False if "conv_layers" in name: load_conv_layer( __snake_case , __snake_case , __snake_case , __snake_case , hf_model.config.feat_extract_norm == '''group''' , ) _UpperCamelCase = True elif name.split('''.''' )[0] == "proj": _UpperCamelCase = fairseq_model.proj _UpperCamelCase = True else: for key, mapped_key in MAPPING.items(): if key in name or key.split('''w2v_model.''' )[-1] == name.split('''.''' )[0]: _UpperCamelCase = True if "*" in mapped_key: _UpperCamelCase = name.split(__snake_case )[0].split('''.''' )[-2] _UpperCamelCase = mapped_key.replace('''*''' , __snake_case ) if "weight_g" in name: _UpperCamelCase = '''weight_g''' elif "weight_v" in name: _UpperCamelCase = '''weight_v''' elif "bias" in name: _UpperCamelCase = '''bias''' elif "weight" in name: _UpperCamelCase = '''weight''' else: _UpperCamelCase = None set_recursively(__snake_case , __snake_case , __snake_case , __snake_case , __snake_case ) continue if not is_used: unused_weights.append(__snake_case ) logger.warning(f"""Unused weights: {unused_weights}""" ) return proj_weight def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = full_name.split('''conv_layers.''' )[-1] _UpperCamelCase = name.split('''.''' ) _UpperCamelCase = int(items[0] ) _UpperCamelCase = int(items[1] ) if type_id == 0: if "bias" in name: assert value.shape == feature_extractor.conv_layers[layer_id].conv.bias.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor.conv_layers[layer_id].conv.bias.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract conv layer {layer_id} was initialized from {full_name}.""" ) elif "weight" in name: assert value.shape == feature_extractor.conv_layers[layer_id].conv.weight.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor.conv_layers[layer_id].conv.weight.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract conv layer {layer_id} was initialized from {full_name}.""" ) elif (type_id == 2 and not use_group_norm) or (type_id == 2 and layer_id == 0 and use_group_norm): if "bias" in name: assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.bias.data.shape, ( f"""{full_name} has size {value.shape}, but {feature_extractor[layer_id].layer_norm.bias.data.shape} was""" " found." ) _UpperCamelCase = value logger.info(f"""Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.""" ) elif "weight" in name: assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.weight.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor[layer_id].layer_norm.weight.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.""" ) else: unused_weights.append(__snake_case ) def _snake_case ( __snake_case ): _UpperCamelCase , _UpperCamelCase = emb.weight.shape _UpperCamelCase = nn.Linear(__snake_case , __snake_case , bias=__snake_case ) _UpperCamelCase = emb.weight.data return lin_layer def _snake_case ( __snake_case ): with open(__snake_case , '''r''' , encoding='''utf-8''' ) as f: _UpperCamelCase = f.readlines() _UpperCamelCase = [line.split(''' ''' )[0] for line in lines] _UpperCamelCase = len(__snake_case ) _UpperCamelCase = { '''<s>''': 0, '''<pad>''': 1, '''</s>''': 2, '''<unk>''': 3, } vocab_dict.update(dict(zip(__snake_case , range(4 , num_words + 4 ) ) ) ) return vocab_dict @torch.no_grad() def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , ): _UpperCamelCase = WavaVecaConfig.from_pretrained(__snake_case ) _UpperCamelCase = SpeechaTextaConfig.from_pretrained( __snake_case , vocab_size=__snake_case , decoder_layers=__snake_case , do_stable_layer_norm=__snake_case ) _UpperCamelCase = WavaVecaFeatureExtractor( feature_size=1 , sampling_rate=16000 , padding_value=0 , do_normalize=__snake_case , return_attention_mask=__snake_case , ) _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = fairseq.checkpoint_utils.load_model_ensemble_and_task( [checkpoint_path] , arg_overrides={'''data''': '''/'''.join(dict_path.split('''/''' )[:-1] )} ) _UpperCamelCase = model[0].eval() # set weights for wav2vec2 encoder _UpperCamelCase = WavaVecaModel(__snake_case ) _UpperCamelCase = recursively_load_weights_wavaveca(model.encoder , __snake_case ) _UpperCamelCase = SpeechaTextaForCausalLM(__snake_case ) _UpperCamelCase , _UpperCamelCase = hf_decoder.model.decoder.load_state_dict(model.decoder.state_dict() , strict=__snake_case ) # set output linear layer unexpected_keys.remove('''embed_out''' ) _UpperCamelCase = nn.Parameter(model.decoder.embed_out.detach() ) # layer norm is init to identity matrix so leaving it is fine logger.warning(f"""The following keys are missing when loading the decoder weights: {missing_keys}""" ) logger.warning(f"""The following keys are unexpected when loading the decoder weights: {unexpected_keys}""" ) _UpperCamelCase = SpeechEncoderDecoderModel(encoder=__snake_case , decoder=__snake_case ) _UpperCamelCase = False # add projection layer _UpperCamelCase = nn.Parameter(projection_layer.weight ) _UpperCamelCase = nn.Parameter(projection_layer.bias ) _UpperCamelCase = create_vocab_dict(__snake_case ) with open(os.path.join(__snake_case , '''vocab.json''' ) , '''w''' ) as fp: json.dump(__snake_case , __snake_case ) _UpperCamelCase = SpeechaTextaTokenizer(os.path.join(__snake_case , '''vocab.json''' ) ) tokenizer.save_pretrained(__snake_case ) _UpperCamelCase = hf_wavavec.config.to_dict() _UpperCamelCase = tokenizer.pad_token_id _UpperCamelCase = tokenizer.bos_token_id _UpperCamelCase = tokenizer.eos_token_id _UpperCamelCase = '''speech_to_text_2''' _UpperCamelCase = '''wav2vec2''' _UpperCamelCase = SpeechEncoderDecoderConfig.from_dict(__snake_case ) hf_wavavec.save_pretrained(__snake_case ) feature_extractor.save_pretrained(__snake_case ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() parser.add_argument("--pytorch_dump_folder_path", default=None, type=str, help="Path to the output PyTorch model.") parser.add_argument("--checkpoint_path", default=None, type=str, help="Path to fairseq checkpoint") parser.add_argument("--dict_path", default=None, type=str, help="Path to dict of fine-tuned model") parser.add_argument( "--encoder_config_path", default="facebook/wav2vec2-large-lv60", type=str, help="Path to hf encoder wav2vec2 checkpoint config", ) parser.add_argument( "--decoder_config_path", default="facebook/s2t-small-mustc-en-fr-st", type=str, help="Path to hf decoder s2t checkpoint config", ) parser.add_argument("--vocab_size", default=10_224, type=int, help="Vocab size of decoder") parser.add_argument("--num_decoder_layers", default=7, type=int, help="Number of decoder layers") _lowerCAmelCase = parser.parse_args() convert_wavaveca_checkpoint( args.checkpoint_path, args.pytorch_dump_folder_path, args.dict_path, encoder_config_path=args.encoder_config_path, decoder_config_path=args.decoder_config_path, vocab_size=args.vocab_size, num_decoder_layers=args.num_decoder_layers, )
71
1
def _snake_case ( __snake_case ): return sum(i for i in range(1 , number // 2 + 1 ) if number % i == 0 ) == number if __name__ == "__main__": print("Program to check whether a number is a Perfect number or not...") _lowerCAmelCase = int(input("Enter number: ").strip()) print(f'{number} is {"" if perfect(number) else "not "}a Perfect Number.')
71
from __future__ import annotations import unittest from transformers import DebertaVaConfig, is_tf_available from transformers.testing_utils import require_tf, slow from ...test_configuration_common import ConfigTester from ...test_modeling_tf_common import TFModelTesterMixin, ids_tensor, random_attention_mask from ...test_pipeline_mixin import PipelineTesterMixin if is_tf_available(): import tensorflow as tf from transformers import ( TFDebertaVaForMaskedLM, TFDebertaVaForQuestionAnswering, TFDebertaVaForSequenceClassification, TFDebertaVaForTokenClassification, TFDebertaVaModel, ) class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : Optional[Any] , _A : List[str]=13 , _A : Union[str, Any]=7 , _A : int=True , _A : Optional[int]=True , _A : Optional[int]=True , _A : Union[str, Any]=True , _A : Optional[int]=99 , _A : Union[str, Any]=32 , _A : Dict=2 , _A : List[Any]=4 , _A : Optional[Any]=37 , _A : int="gelu" , _A : Optional[int]=0.1 , _A : str=0.1 , _A : List[str]=512 , _A : Optional[Any]=16 , _A : Optional[Any]=2 , _A : Optional[int]=0.02 , _A : str=False , _A : int=True , _A : Any="None" , _A : Dict=3 , _A : List[Any]=4 , _A : Optional[Any]=None , ): _UpperCamelCase = parent _UpperCamelCase = batch_size _UpperCamelCase = seq_length _UpperCamelCase = is_training _UpperCamelCase = use_input_mask _UpperCamelCase = use_token_type_ids _UpperCamelCase = use_labels _UpperCamelCase = vocab_size _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = hidden_dropout_prob _UpperCamelCase = attention_probs_dropout_prob _UpperCamelCase = max_position_embeddings _UpperCamelCase = type_vocab_size _UpperCamelCase = type_sequence_label_size _UpperCamelCase = initializer_range _UpperCamelCase = num_labels _UpperCamelCase = num_choices _UpperCamelCase = relative_attention _UpperCamelCase = position_biased_input _UpperCamelCase = pos_att_type _UpperCamelCase = scope def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size ) _UpperCamelCase = None if self.use_input_mask: _UpperCamelCase = random_attention_mask([self.batch_size, self.seq_length] ) _UpperCamelCase = None if self.use_token_type_ids: _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.type_vocab_size ) _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None if self.use_labels: _UpperCamelCase = ids_tensor([self.batch_size] , self.type_sequence_label_size ) _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.num_labels ) _UpperCamelCase = DebertaVaConfig( vocab_size=self.vocab_size , hidden_size=self.hidden_size , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , relative_attention=self.relative_attention , position_biased_input=self.position_biased_input , initializer_range=self.initializer_range , return_dict=_A , ) return config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels def UpperCamelCase_ ( self : Dict , _A : Tuple , _A : Tuple , _A : Union[str, Any] , _A : List[str] , _A : Optional[int] , _A : int , _A : Optional[Any] ): _UpperCamelCase = TFDebertaVaModel(config=_A ) _UpperCamelCase = {'''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids} _UpperCamelCase = [input_ids, input_mask] _UpperCamelCase = model(_A ) _UpperCamelCase = model(_A ) self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) ) def UpperCamelCase_ ( self : Dict , _A : Optional[int] , _A : Any , _A : Dict , _A : Union[str, Any] , _A : Union[str, Any] , _A : List[Any] , _A : List[str] ): _UpperCamelCase = TFDebertaVaForMaskedLM(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) ) def UpperCamelCase_ ( self : Dict , _A : Dict , _A : List[str] , _A : List[Any] , _A : List[Any] , _A : Optional[Any] , _A : Tuple , _A : int ): _UpperCamelCase = self.num_labels _UpperCamelCase = TFDebertaVaForSequenceClassification(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) ) def UpperCamelCase_ ( self : Tuple , _A : Dict , _A : Optional[int] , _A : Any , _A : List[Any] , _A : Dict , _A : Union[str, Any] , _A : List[str] ): _UpperCamelCase = self.num_labels _UpperCamelCase = TFDebertaVaForTokenClassification(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) ) def UpperCamelCase_ ( self : Dict , _A : Optional[Any] , _A : Optional[int] , _A : Any , _A : List[str] , _A : str , _A : Optional[int] , _A : str ): _UpperCamelCase = TFDebertaVaForQuestionAnswering(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.start_logits.shape , (self.batch_size, self.seq_length) ) self.parent.assertEqual(result.end_logits.shape , (self.batch_size, self.seq_length) ) def UpperCamelCase_ ( self : Any ): _UpperCamelCase = self.prepare_config_and_inputs() ( ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ) = config_and_inputs _UpperCamelCase = {'''input_ids''': input_ids, '''token_type_ids''': token_type_ids, '''attention_mask''': input_mask} return config, inputs_dict @require_tf class lowerCAmelCase_ ( __lowercase, __lowercase, unittest.TestCase ): UpperCAmelCase = ( ( TFDebertaVaModel, TFDebertaVaForMaskedLM, TFDebertaVaForQuestionAnswering, TFDebertaVaForSequenceClassification, TFDebertaVaForTokenClassification, ) if is_tf_available() else () ) UpperCAmelCase = ( { "feature-extraction": TFDebertaVaModel, "fill-mask": TFDebertaVaForMaskedLM, "question-answering": TFDebertaVaForQuestionAnswering, "text-classification": TFDebertaVaForSequenceClassification, "token-classification": TFDebertaVaForTokenClassification, "zero-shot": TFDebertaVaForSequenceClassification, } if is_tf_available() else {} ) UpperCAmelCase = False UpperCAmelCase = False def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = TFDebertaVaModelTester(self ) _UpperCamelCase = ConfigTester(self , config_class=_A , hidden_size=37 ) def UpperCamelCase_ ( self : Any ): self.config_tester.run_common_tests() def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model(*_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_masked_lm(*_A ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_question_answering(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_sequence_classification(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_token_classification(*_A ) @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = TFDebertaVaModel.from_pretrained('''kamalkraj/deberta-v2-xlarge''' ) self.assertIsNotNone(_A ) @require_tf class lowerCAmelCase_ ( unittest.TestCase ): @unittest.skip(reason='''Model not available yet''' ) def UpperCamelCase_ ( self : List[Any] ): pass @slow def UpperCamelCase_ ( self : int ): _UpperCamelCase = TFDebertaVaModel.from_pretrained('''kamalkraj/deberta-v2-xlarge''' ) _UpperCamelCase = tf.constant([[0, 3_1414, 232, 328, 740, 1140, 1_2695, 69, 4_6078, 1588, 2]] ) _UpperCamelCase = tf.constant([[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]] ) _UpperCamelCase = model(_A , attention_mask=_A )[0] _UpperCamelCase = tf.constant( [[[0.2356, 0.1948, 0.0369], [-0.1063, 0.3586, -0.5152], [-0.6399, -0.0259, -0.2525]]] ) tf.debugging.assert_near(output[:, 1:4, 1:4] , _A , atol=1e-4 )
71
1
from typing import Optional from torch import nn from .transformer_ad import TransformeraDModel, TransformeraDModelOutput class lowerCAmelCase_ ( nn.Module ): def __init__( self : Optional[int] , _A : int = 16 , _A : int = 88 , _A : Optional[int] = None , _A : int = 1 , _A : float = 0.0 , _A : int = 32 , _A : Optional[int] = None , _A : bool = False , _A : Optional[int] = None , _A : Optional[int] = None , _A : str = "geglu" , _A : Optional[int] = None , ): super().__init__() _UpperCamelCase = nn.ModuleList( [ TransformeraDModel( num_attention_heads=_A , attention_head_dim=_A , in_channels=_A , num_layers=_A , dropout=_A , norm_num_groups=_A , cross_attention_dim=_A , attention_bias=_A , sample_size=_A , num_vector_embeds=_A , activation_fn=_A , num_embeds_ada_norm=_A , ) for _ in range(2 ) ] ) # Variables that can be set by a pipeline: # The ratio of transformer1 to transformer2's output states to be combined during inference _UpperCamelCase = 0.5 # The shape of `encoder_hidden_states` is expected to be # `(batch_size, condition_lengths[0]+condition_lengths[1], num_features)` _UpperCamelCase = [77, 257] # Which transformer to use to encode which condition. # E.g. `(1, 0)` means that we'll use `transformers[1](conditions[0])` and `transformers[0](conditions[1])` _UpperCamelCase = [1, 0] def UpperCamelCase_ ( self : int , _A : Any , _A : int , _A : Tuple=None , _A : Union[str, Any]=None , _A : Tuple=None , _A : bool = True , ): _UpperCamelCase = hidden_states _UpperCamelCase = [] _UpperCamelCase = 0 # attention_mask is not used yet for i in range(2 ): # for each of the two transformers, pass the corresponding condition tokens _UpperCamelCase = encoder_hidden_states[:, tokens_start : tokens_start + self.condition_lengths[i]] _UpperCamelCase = self.transformer_index_for_condition[i] _UpperCamelCase = self.transformers[transformer_index]( _A , encoder_hidden_states=_A , timestep=_A , cross_attention_kwargs=_A , return_dict=_A , )[0] encoded_states.append(encoded_state - input_states ) tokens_start += self.condition_lengths[i] _UpperCamelCase = encoded_states[0] * self.mix_ratio + encoded_states[1] * (1 - self.mix_ratio) _UpperCamelCase = output_states + input_states if not return_dict: return (output_states,) return TransformeraDModelOutput(sample=_A )
71
def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): # Return True if there is node that has not iterated. _UpperCamelCase = [False] * len(__snake_case ) _UpperCamelCase = [] queue.append(__snake_case ) _UpperCamelCase = True while queue: _UpperCamelCase = queue.pop(0 ) for ind in range(len(graph[u] ) ): if visited[ind] is False and graph[u][ind] > 0: queue.append(__snake_case ) _UpperCamelCase = True _UpperCamelCase = u return visited[t] def _snake_case ( __snake_case , __snake_case , __snake_case ): # This array is filled by BFS and to store path _UpperCamelCase = [-1] * (len(__snake_case )) _UpperCamelCase = 0 while bfs(__snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = float('''Inf''' ) _UpperCamelCase = sink while s != source: # Find the minimum value in select path _UpperCamelCase = min(__snake_case , graph[parent[s]][s] ) _UpperCamelCase = parent[s] max_flow += path_flow _UpperCamelCase = sink while v != source: _UpperCamelCase = parent[v] graph[u][v] -= path_flow graph[v][u] += path_flow _UpperCamelCase = parent[v] return max_flow _lowerCAmelCase = [ [0, 16, 13, 0, 0, 0], [0, 0, 10, 12, 0, 0], [0, 4, 0, 0, 14, 0], [0, 0, 9, 0, 0, 20], [0, 0, 0, 7, 0, 4], [0, 0, 0, 0, 0, 0], ] _lowerCAmelCase, _lowerCAmelCase = 0, 5 print(ford_fulkerson(graph, source, sink))
71
1
import argparse import json import os import torch from transformers import LukeConfig, LukeModel, LukeTokenizer, RobertaTokenizer from transformers.tokenization_utils_base import AddedToken @torch.no_grad() def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): # Load configuration defined in the metadata file with open(__snake_case ) as metadata_file: _UpperCamelCase = json.load(__snake_case ) _UpperCamelCase = LukeConfig(use_entity_aware_attention=__snake_case , **metadata['''model_config'''] ) # Load in the weights from the checkpoint_path _UpperCamelCase = torch.load(__snake_case , map_location='''cpu''' ) # Load the entity vocab file _UpperCamelCase = load_entity_vocab(__snake_case ) _UpperCamelCase = RobertaTokenizer.from_pretrained(metadata['''model_config''']['''bert_model_name'''] ) # Add special tokens to the token vocabulary for downstream tasks _UpperCamelCase = AddedToken('''<ent>''' , lstrip=__snake_case , rstrip=__snake_case ) _UpperCamelCase = AddedToken('''<ent2>''' , lstrip=__snake_case , rstrip=__snake_case ) tokenizer.add_special_tokens({'''additional_special_tokens''': [entity_token_a, entity_token_a]} ) config.vocab_size += 2 print(f"""Saving tokenizer to {pytorch_dump_folder_path}""" ) tokenizer.save_pretrained(__snake_case ) with open(os.path.join(__snake_case , LukeTokenizer.vocab_files_names['''entity_vocab_file'''] ) , '''w''' ) as f: json.dump(__snake_case , __snake_case ) _UpperCamelCase = LukeTokenizer.from_pretrained(__snake_case ) # Initialize the embeddings of the special tokens _UpperCamelCase = state_dict['''embeddings.word_embeddings.weight'''] _UpperCamelCase = word_emb[tokenizer.convert_tokens_to_ids(['''@'''] )[0]].unsqueeze(0 ) _UpperCamelCase = word_emb[tokenizer.convert_tokens_to_ids(['''#'''] )[0]].unsqueeze(0 ) _UpperCamelCase = torch.cat([word_emb, ent_emb, enta_emb] ) # Initialize the query layers of the entity-aware self-attention mechanism for layer_index in range(config.num_hidden_layers ): for matrix_name in ["query.weight", "query.bias"]: _UpperCamelCase = f"""encoder.layer.{layer_index}.attention.self.""" _UpperCamelCase = state_dict[prefix + matrix_name] _UpperCamelCase = state_dict[prefix + matrix_name] _UpperCamelCase = state_dict[prefix + matrix_name] # Initialize the embedding of the [MASK2] entity using that of the [MASK] entity for downstream tasks _UpperCamelCase = state_dict['''entity_embeddings.entity_embeddings.weight'''] _UpperCamelCase = entity_emb[entity_vocab['''[MASK]''']] _UpperCamelCase = LukeModel(config=__snake_case ).eval() _UpperCamelCase , _UpperCamelCase = model.load_state_dict(__snake_case , strict=__snake_case ) if not (len(__snake_case ) == 1 and missing_keys[0] == "embeddings.position_ids"): raise ValueError(f"""Missing keys {", ".join(__snake_case )}. Expected only missing embeddings.position_ids""" ) if not (all(key.startswith('''entity_predictions''' ) or key.startswith('''lm_head''' ) for key in unexpected_keys )): raise ValueError( '''Unexpected keys''' f""" {", ".join([key for key in unexpected_keys if not (key.startswith("entity_predictions" ) or key.startswith("lm_head" ))] )}""" ) # Check outputs _UpperCamelCase = LukeTokenizer.from_pretrained(__snake_case , task='''entity_classification''' ) _UpperCamelCase = ( '''Top seed Ana Ivanovic said on Thursday she could hardly believe her luck as a fortuitous netcord helped the''' ''' new world number one avoid a humiliating second- round exit at Wimbledon .''' ) _UpperCamelCase = (39, 42) _UpperCamelCase = tokenizer(__snake_case , entity_spans=[span] , add_prefix_space=__snake_case , return_tensors='''pt''' ) _UpperCamelCase = model(**__snake_case ) # Verify word hidden states if model_size == "large": _UpperCamelCase = torch.Size((1, 42, 1024) ) _UpperCamelCase = torch.tensor( [[0.0133, 0.0865, 0.0095], [0.3093, -0.2576, -0.7418], [-0.1720, -0.2117, -0.2869]] ) else: # base _UpperCamelCase = torch.Size((1, 42, 768) ) _UpperCamelCase = torch.tensor([[0.0037, 0.1368, -0.0091], [0.1099, 0.3329, -0.1095], [0.0765, 0.5335, 0.1179]] ) if not (outputs.last_hidden_state.shape == expected_shape): raise ValueError( f"""Outputs.last_hidden_state.shape is {outputs.last_hidden_state.shape}, Expected shape is {expected_shape}""" ) if not torch.allclose(outputs.last_hidden_state[0, :3, :3] , __snake_case , atol=1E-4 ): raise ValueError # Verify entity hidden states if model_size == "large": _UpperCamelCase = torch.Size((1, 1, 1024) ) _UpperCamelCase = torch.tensor([[0.0466, -0.0106, -0.0179]] ) else: # base _UpperCamelCase = torch.Size((1, 1, 768) ) _UpperCamelCase = torch.tensor([[0.1457, 0.1044, 0.0174]] ) if not (outputs.entity_last_hidden_state.shape != expected_shape): raise ValueError( f"""Outputs.entity_last_hidden_state.shape is {outputs.entity_last_hidden_state.shape}, Expected shape is""" f""" {expected_shape}""" ) if not torch.allclose(outputs.entity_last_hidden_state[0, :3, :3] , __snake_case , atol=1E-4 ): raise ValueError # Finally, save our PyTorch model and tokenizer print('''Saving PyTorch model to {}'''.format(__snake_case ) ) model.save_pretrained(__snake_case ) def _snake_case ( __snake_case ): _UpperCamelCase = {} with open(__snake_case , '''r''' , encoding='''utf-8''' ) as f: for index, line in enumerate(__snake_case ): _UpperCamelCase , _UpperCamelCase = line.rstrip().split('''\t''' ) _UpperCamelCase = index return entity_vocab if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() # Required parameters parser.add_argument("--checkpoint_path", type=str, help="Path to a pytorch_model.bin file.") parser.add_argument( "--metadata_path", default=None, type=str, help="Path to a metadata.json file, defining the configuration." ) parser.add_argument( "--entity_vocab_path", default=None, type=str, help="Path to an entity_vocab.tsv file, containing the entity vocabulary.", ) parser.add_argument( "--pytorch_dump_folder_path", default=None, type=str, help="Path to where to dump the output PyTorch model." ) parser.add_argument( "--model_size", default="base", type=str, choices=["base", "large"], help="Size of the model to be converted." ) _lowerCAmelCase = parser.parse_args() convert_luke_checkpoint( args.checkpoint_path, args.metadata_path, args.entity_vocab_path, args.pytorch_dump_folder_path, args.model_size, )
71
from typing import TYPE_CHECKING from ...utils import ( OptionalDependencyNotAvailable, _LazyModule, is_flax_available, is_tf_available, is_torch_available, ) _lowerCAmelCase = {"configuration_unispeech": ["UNISPEECH_PRETRAINED_CONFIG_ARCHIVE_MAP", "UniSpeechConfig"]} try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "UNISPEECH_PRETRAINED_MODEL_ARCHIVE_LIST", "UniSpeechForCTC", "UniSpeechForPreTraining", "UniSpeechForSequenceClassification", "UniSpeechModel", "UniSpeechPreTrainedModel", ] if TYPE_CHECKING: from .configuration_unispeech import UNISPEECH_PRETRAINED_CONFIG_ARCHIVE_MAP, UniSpeechConfig try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_unispeech import ( UNISPEECH_PRETRAINED_MODEL_ARCHIVE_LIST, UniSpeechForCTC, UniSpeechForPreTraining, UniSpeechForSequenceClassification, UniSpeechModel, UniSpeechPreTrainedModel, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
1
import copy import os from collections import OrderedDict from typing import TYPE_CHECKING, Any, Dict, Mapping, Optional, Union if TYPE_CHECKING: from ...processing_utils import ProcessorMixin from ...utils import TensorType from ...configuration_utils import PretrainedConfig from ...onnx import OnnxConfig from ...utils import logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "google/owlvit-base-patch32": "https://huggingface.co/google/owlvit-base-patch32/resolve/main/config.json", "google/owlvit-base-patch16": "https://huggingface.co/google/owlvit-base-patch16/resolve/main/config.json", "google/owlvit-large-patch14": "https://huggingface.co/google/owlvit-large-patch14/resolve/main/config.json", } class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "owlvit_text_model" def __init__( self : Optional[int] , _A : Union[str, Any]=4_9408 , _A : Optional[int]=512 , _A : Dict=2048 , _A : Tuple=12 , _A : Optional[int]=8 , _A : str=16 , _A : Tuple="quick_gelu" , _A : List[str]=1e-5 , _A : Optional[Any]=0.0 , _A : int=0.02 , _A : List[str]=1.0 , _A : Tuple=0 , _A : Any=4_9406 , _A : Any=4_9407 , **_A : Optional[Any] , ): super().__init__(pad_token_id=_A , bos_token_id=_A , eos_token_id=_A , **_A ) _UpperCamelCase = vocab_size _UpperCamelCase = hidden_size _UpperCamelCase = intermediate_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = max_position_embeddings _UpperCamelCase = hidden_act _UpperCamelCase = layer_norm_eps _UpperCamelCase = attention_dropout _UpperCamelCase = initializer_range _UpperCamelCase = initializer_factor @classmethod def UpperCamelCase_ ( cls : str , _A : Union[str, os.PathLike] , **_A : List[Any] ): cls._set_token_in_kwargs(_A ) _UpperCamelCase , _UpperCamelCase = cls.get_config_dict(_A , **_A ) # get the text config dict if we are loading from OwlViTConfig if config_dict.get('''model_type''' ) == "owlvit": _UpperCamelCase = config_dict['''text_config'''] if "model_type" in config_dict and hasattr(cls , '''model_type''' ) and config_dict["model_type"] != cls.model_type: logger.warning( F"""You are using a model of type {config_dict["model_type"]} to instantiate a model of type """ F"""{cls.model_type}. This is not supported for all configurations of models and can yield errors.""" ) return cls.from_dict(_A , **_A ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "owlvit_vision_model" def __init__( self : List[Any] , _A : Optional[int]=768 , _A : List[Any]=3072 , _A : List[Any]=12 , _A : Any=12 , _A : Union[str, Any]=3 , _A : List[str]=768 , _A : Union[str, Any]=32 , _A : Union[str, Any]="quick_gelu" , _A : str=1e-5 , _A : int=0.0 , _A : Any=0.02 , _A : Dict=1.0 , **_A : Optional[int] , ): super().__init__(**_A ) _UpperCamelCase = hidden_size _UpperCamelCase = intermediate_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = num_channels _UpperCamelCase = image_size _UpperCamelCase = patch_size _UpperCamelCase = hidden_act _UpperCamelCase = layer_norm_eps _UpperCamelCase = attention_dropout _UpperCamelCase = initializer_range _UpperCamelCase = initializer_factor @classmethod def UpperCamelCase_ ( cls : Tuple , _A : Union[str, os.PathLike] , **_A : List[Any] ): cls._set_token_in_kwargs(_A ) _UpperCamelCase , _UpperCamelCase = cls.get_config_dict(_A , **_A ) # get the vision config dict if we are loading from OwlViTConfig if config_dict.get('''model_type''' ) == "owlvit": _UpperCamelCase = config_dict['''vision_config'''] if "model_type" in config_dict and hasattr(cls , '''model_type''' ) and config_dict["model_type"] != cls.model_type: logger.warning( F"""You are using a model of type {config_dict["model_type"]} to instantiate a model of type """ F"""{cls.model_type}. This is not supported for all configurations of models and can yield errors.""" ) return cls.from_dict(_A , **_A ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "owlvit" UpperCAmelCase = True def __init__( self : Dict , _A : List[Any]=None , _A : Union[str, Any]=None , _A : Any=512 , _A : Any=2.6592 , _A : List[Any]=True , **_A : int , ): super().__init__(**_A ) if text_config is None: _UpperCamelCase = {} logger.info('''text_config is None. Initializing the OwlViTTextConfig with default values.''' ) if vision_config is None: _UpperCamelCase = {} logger.info('''vision_config is None. initializing the OwlViTVisionConfig with default values.''' ) _UpperCamelCase = OwlViTTextConfig(**_A ) _UpperCamelCase = OwlViTVisionConfig(**_A ) _UpperCamelCase = projection_dim _UpperCamelCase = logit_scale_init_value _UpperCamelCase = return_dict _UpperCamelCase = 1.0 @classmethod def UpperCamelCase_ ( cls : Union[str, Any] , _A : Union[str, os.PathLike] , **_A : Optional[Any] ): cls._set_token_in_kwargs(_A ) _UpperCamelCase , _UpperCamelCase = cls.get_config_dict(_A , **_A ) if "model_type" in config_dict and hasattr(cls , '''model_type''' ) and config_dict["model_type"] != cls.model_type: logger.warning( F"""You are using a model of type {config_dict["model_type"]} to instantiate a model of type """ F"""{cls.model_type}. This is not supported for all configurations of models and can yield errors.""" ) return cls.from_dict(_A , **_A ) @classmethod def UpperCamelCase_ ( cls : Union[str, Any] , _A : Dict , _A : Dict , **_A : Optional[int] ): _UpperCamelCase = {} _UpperCamelCase = text_config _UpperCamelCase = vision_config return cls.from_dict(_A , **_A ) def UpperCamelCase_ ( self : str ): _UpperCamelCase = copy.deepcopy(self.__dict__ ) _UpperCamelCase = self.text_config.to_dict() _UpperCamelCase = self.vision_config.to_dict() _UpperCamelCase = self.__class__.model_type return output class lowerCAmelCase_ ( __lowercase ): @property def UpperCamelCase_ ( self : Optional[Any] ): return OrderedDict( [ ('''input_ids''', {0: '''batch''', 1: '''sequence'''}), ('''pixel_values''', {0: '''batch''', 1: '''num_channels''', 2: '''height''', 3: '''width'''}), ('''attention_mask''', {0: '''batch''', 1: '''sequence'''}), ] ) @property def UpperCamelCase_ ( self : str ): return OrderedDict( [ ('''logits_per_image''', {0: '''batch'''}), ('''logits_per_text''', {0: '''batch'''}), ('''text_embeds''', {0: '''batch'''}), ('''image_embeds''', {0: '''batch'''}), ] ) @property def UpperCamelCase_ ( self : int ): return 1e-4 def UpperCamelCase_ ( self : List[Any] , _A : "ProcessorMixin" , _A : int = -1 , _A : int = -1 , _A : Optional["TensorType"] = None , ): _UpperCamelCase = super().generate_dummy_inputs( processor.tokenizer , batch_size=_A , seq_length=_A , framework=_A ) _UpperCamelCase = super().generate_dummy_inputs( processor.image_processor , batch_size=_A , framework=_A ) return {**text_input_dict, **image_input_dict} @property def UpperCamelCase_ ( self : Dict ): return 14
71
import json import os import shutil import tempfile import unittest import numpy as np from transformers import BertTokenizerFast from transformers.models.bert.tokenization_bert import VOCAB_FILES_NAMES, BertTokenizer from transformers.testing_utils import require_tokenizers, require_vision from transformers.utils import IMAGE_PROCESSOR_NAME, is_vision_available if is_vision_available(): from PIL import Image from transformers import VisionTextDualEncoderProcessor, ViTImageProcessor @require_tokenizers @require_vision class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = tempfile.mkdtemp() # fmt: off _UpperCamelCase = ['''[UNK]''', '''[CLS]''', '''[SEP]''', '''[PAD]''', '''[MASK]''', '''want''', '''##want''', '''##ed''', '''wa''', '''un''', '''runn''', '''##ing''', ''',''', '''low''', '''lowest'''] # fmt: on _UpperCamelCase = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES['''vocab_file'''] ) with open(self.vocab_file , '''w''' , encoding='''utf-8''' ) as vocab_writer: vocab_writer.write(''''''.join([x + '''\n''' for x in vocab_tokens] ) ) _UpperCamelCase = { '''do_resize''': True, '''size''': {'''height''': 18, '''width''': 18}, '''do_normalize''': True, '''image_mean''': [0.5, 0.5, 0.5], '''image_std''': [0.5, 0.5, 0.5], } _UpperCamelCase = os.path.join(self.tmpdirname , _A ) with open(self.image_processor_file , '''w''' , encoding='''utf-8''' ) as fp: json.dump(_A , _A ) def UpperCamelCase_ ( self : Tuple , **_A : Optional[Any] ): return BertTokenizer.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : List[Any] , **_A : Union[str, Any] ): return ViTImageProcessor.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : int ): shutil.rmtree(self.tmpdirname ) def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = [np.random.randint(255 , size=(3, 30, 400) , dtype=np.uinta )] _UpperCamelCase = [Image.fromarray(np.moveaxis(_A , 0 , -1 ) ) for x in image_inputs] return image_inputs def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = self.get_image_processor() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) processor.save_pretrained(self.tmpdirname ) _UpperCamelCase = VisionTextDualEncoderProcessor.from_pretrained(self.tmpdirname ) self.assertEqual(processor.tokenizer.get_vocab() , tokenizer.get_vocab() ) self.assertIsInstance(processor.tokenizer , (BertTokenizer, BertTokenizerFast) ) self.assertEqual(processor.image_processor.to_json_string() , image_processor.to_json_string() ) self.assertIsInstance(processor.image_processor , _A ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = VisionTextDualEncoderProcessor( tokenizer=self.get_tokenizer() , image_processor=self.get_image_processor() ) processor.save_pretrained(self.tmpdirname ) _UpperCamelCase = self.get_tokenizer(bos_token='''(BOS)''' , eos_token='''(EOS)''' ) _UpperCamelCase = self.get_image_processor(do_normalize=_A , padding_value=1.0 ) _UpperCamelCase = VisionTextDualEncoderProcessor.from_pretrained( self.tmpdirname , bos_token='''(BOS)''' , eos_token='''(EOS)''' , do_normalize=_A , padding_value=1.0 ) self.assertEqual(processor.tokenizer.get_vocab() , tokenizer_add_kwargs.get_vocab() ) self.assertIsInstance(processor.tokenizer , (BertTokenizer, BertTokenizerFast) ) self.assertEqual(processor.image_processor.to_json_string() , image_processor_add_kwargs.to_json_string() ) self.assertIsInstance(processor.image_processor , _A ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = image_processor(_A , return_tensors='''np''' ) _UpperCamelCase = processor(images=_A , return_tensors='''np''' ) for key in input_feat_extract.keys(): self.assertAlmostEqual(input_feat_extract[key].sum() , input_processor[key].sum() , delta=1e-2 ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = processor(text=_A ) _UpperCamelCase = tokenizer(_A ) for key in encoded_tok.keys(): self.assertListEqual(encoded_tok[key] , encoded_processor[key] ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = processor(text=_A , images=_A ) self.assertListEqual(list(inputs.keys() ) , ['''input_ids''', '''token_type_ids''', '''attention_mask''', '''pixel_values'''] ) # test if it raises when no input is passed with self.assertRaises(_A ): processor() def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = [[1, 4, 5, 8, 1, 0, 8], [3, 4, 3, 1, 1, 8, 9]] _UpperCamelCase = processor.batch_decode(_A ) _UpperCamelCase = tokenizer.batch_decode(_A ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = processor(text=_A , images=_A ) self.assertListEqual(list(inputs.keys() ) , processor.model_input_names )
71
1
import unittest from dataclasses import dataclass import pytest from accelerate.commands.config.config_args import SageMakerConfig from accelerate.utils import ComputeEnvironment from accelerate.utils.launch import _convert_nargs_to_dict @dataclass class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = ComputeEnvironment.AMAZON_SAGEMAKER UpperCAmelCase = True UpperCAmelCase = "ml.p3.2xlarge" UpperCAmelCase = "accelerate_sagemaker_execution_role" UpperCAmelCase = "hf-sm" UpperCAmelCase = "us-east-1" UpperCAmelCase = 1 UpperCAmelCase = "accelerate-sagemaker-1" UpperCAmelCase = "1.6" UpperCAmelCase = "4.4" UpperCAmelCase = "train.py" UpperCAmelCase = [ "--model_name_or_path", "bert", "--do_train", "False", "--epochs", "3", "--learning_rate", "5e-5", "--max_steps", "50.5", ] UpperCAmelCase = [ "--model_name_or_path", "bert", "--do_train", "--do_test", "False", "--do_predict", "--epochs", "3", "--learning_rate", "5e-5", "--max_steps", "50.5", ] class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : List[Any] ): # If no defaults are changed, `to_kwargs` returns an empty dict. _UpperCamelCase = _convert_nargs_to_dict(MockLaunchConfig.success_training_script_args ) assert isinstance(converted_args['''model_name_or_path'''] , _A ) assert isinstance(converted_args['''do_train'''] , _A ) assert isinstance(converted_args['''epochs'''] , _A ) assert isinstance(converted_args['''learning_rate'''] , _A ) assert isinstance(converted_args['''max_steps'''] , _A ) with pytest.raises(_A ): _convert_nargs_to_dict(MockLaunchConfig.fail_training_script_args )
71
def _snake_case ( __snake_case , __snake_case , __snake_case ): if n == 0: return 1 elif n % 2 == 1: return (binary_exponentiation(__snake_case , n - 1 , __snake_case ) * a) % mod else: _UpperCamelCase = binary_exponentiation(__snake_case , n / 2 , __snake_case ) return (b * b) % mod # a prime number _lowerCAmelCase = 701 _lowerCAmelCase = 1_000_000_000 _lowerCAmelCase = 10 # using binary exponentiation function, O(log(p)): print((a / b) % p == (a * binary_exponentiation(b, p - 2, p)) % p) print((a / b) % p == (a * b ** (p - 2)) % p)
71
1
from __future__ import annotations from collections import deque class lowerCAmelCase_ : def __init__( self : List[str] , _A : list[str] ): _UpperCamelCase = [] self.adlist.append( {'''value''': '''''', '''next_states''': [], '''fail_state''': 0, '''output''': []} ) for keyword in keywords: self.add_keyword(_A ) self.set_fail_transitions() def UpperCamelCase_ ( self : List[Any] , _A : int , _A : str ): for state in self.adlist[current_state]["next_states"]: if char == self.adlist[state]["value"]: return state return None def UpperCamelCase_ ( self : Tuple , _A : str ): _UpperCamelCase = 0 for character in keyword: _UpperCamelCase = self.find_next_state(_A , _A ) if next_state is None: self.adlist.append( { '''value''': character, '''next_states''': [], '''fail_state''': 0, '''output''': [], } ) self.adlist[current_state]["next_states"].append(len(self.adlist ) - 1 ) _UpperCamelCase = len(self.adlist ) - 1 else: _UpperCamelCase = next_state self.adlist[current_state]["output"].append(_A ) def UpperCamelCase_ ( self : Any ): _UpperCamelCase = deque() for node in self.adlist[0]["next_states"]: q.append(_A ) _UpperCamelCase = 0 while q: _UpperCamelCase = q.popleft() for child in self.adlist[r]["next_states"]: q.append(_A ) _UpperCamelCase = self.adlist[r]['''fail_state'''] while ( self.find_next_state(_A , self.adlist[child]['''value'''] ) is None and state != 0 ): _UpperCamelCase = self.adlist[state]['''fail_state'''] _UpperCamelCase = self.find_next_state( _A , self.adlist[child]['''value'''] ) if self.adlist[child]["fail_state"] is None: _UpperCamelCase = 0 _UpperCamelCase = ( self.adlist[child]['''output'''] + self.adlist[self.adlist[child]['''fail_state''']]['''output'''] ) def UpperCamelCase_ ( self : int , _A : str ): _UpperCamelCase = {} # returns a dict with keywords and list of its occurrences _UpperCamelCase = 0 for i in range(len(_A ) ): while ( self.find_next_state(_A , string[i] ) is None and current_state != 0 ): _UpperCamelCase = self.adlist[current_state]['''fail_state'''] _UpperCamelCase = self.find_next_state(_A , string[i] ) if next_state is None: _UpperCamelCase = 0 else: _UpperCamelCase = next_state for key in self.adlist[current_state]["output"]: if key not in result: _UpperCamelCase = [] result[key].append(i - len(_A ) + 1 ) return result if __name__ == "__main__": import doctest doctest.testmod()
71
from math import cos, sin, sqrt, tau from audio_filters.iir_filter import IIRFilter def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 - _cos) / 2 _UpperCamelCase = 1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 + _cos) / 2 _UpperCamelCase = -1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = _sin / 2 _UpperCamelCase = 0 _UpperCamelCase = -ba _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 1 - alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([ba, ba, ba] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = 1 + alpha * big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha * big_a _UpperCamelCase = 1 + alpha / big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha / big_a _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (pmc + aaa) _UpperCamelCase = 2 * big_a * mpc _UpperCamelCase = big_a * (pmc - aaa) _UpperCamelCase = ppmc + aaa _UpperCamelCase = -2 * pmpc _UpperCamelCase = ppmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (ppmc + aaa) _UpperCamelCase = -2 * big_a * pmpc _UpperCamelCase = big_a * (ppmc - aaa) _UpperCamelCase = pmc + aaa _UpperCamelCase = 2 * mpc _UpperCamelCase = pmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt
71
1
import os from shutil import copyfile from typing import Any, Dict, List, Optional, Tuple import sentencepiece as spm from ...tokenization_utils import AddedToken, BatchEncoding, PreTrainedTokenizer from ...utils import logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = "▁" _lowerCAmelCase = {"vocab_file": "sentencepiece.bpe.model"} _lowerCAmelCase = { "vocab_file": { "facebook/mbart-large-50-one-to-many-mmt": ( "https://huggingface.co/facebook/mbart-large-50-one-to-many-mmt/resolve/main/sentencepiece.bpe.model" ), } } _lowerCAmelCase = { "facebook/mbart-large-50-one-to-many-mmt": 1_024, } # fmt: off _lowerCAmelCase = ["ar_AR", "cs_CZ", "de_DE", "en_XX", "es_XX", "et_EE", "fi_FI", "fr_XX", "gu_IN", "hi_IN", "it_IT", "ja_XX", "kk_KZ", "ko_KR", "lt_LT", "lv_LV", "my_MM", "ne_NP", "nl_XX", "ro_RO", "ru_RU", "si_LK", "tr_TR", "vi_VN", "zh_CN", "af_ZA", "az_AZ", "bn_IN", "fa_IR", "he_IL", "hr_HR", "id_ID", "ka_GE", "km_KH", "mk_MK", "ml_IN", "mn_MN", "mr_IN", "pl_PL", "ps_AF", "pt_XX", "sv_SE", "sw_KE", "ta_IN", "te_IN", "th_TH", "tl_XX", "uk_UA", "ur_PK", "xh_ZA", "gl_ES", "sl_SI"] class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = VOCAB_FILES_NAMES UpperCAmelCase = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES UpperCAmelCase = PRETRAINED_VOCAB_FILES_MAP UpperCAmelCase = ["input_ids", "attention_mask"] UpperCAmelCase = [] UpperCAmelCase = [] def __init__( self : Tuple , _A : str , _A : Any=None , _A : List[Any]=None , _A : List[str]="</s>" , _A : Optional[int]="</s>" , _A : List[Any]="<s>" , _A : int="<unk>" , _A : List[Any]="<pad>" , _A : List[str]="<mask>" , _A : Optional[Dict[str, Any]] = None , **_A : List[Any] , ): # Mask token behave like a normal word, i.e. include the space before it _UpperCamelCase = AddedToken(_A , lstrip=_A , rstrip=_A ) if isinstance(_A , _A ) else mask_token _UpperCamelCase = {} if sp_model_kwargs is None else sp_model_kwargs _UpperCamelCase = kwargs.get('''additional_special_tokens''' , [] ) kwargs["additional_special_tokens"] += [ code for code in FAIRSEQ_LANGUAGE_CODES if code not in kwargs["additional_special_tokens"] ] super().__init__( src_lang=_A , tgt_lang=_A , eos_token=_A , unk_token=_A , sep_token=_A , cls_token=_A , pad_token=_A , mask_token=_A , sp_model_kwargs=self.sp_model_kwargs , **_A , ) _UpperCamelCase = spm.SentencePieceProcessor(**self.sp_model_kwargs ) self.sp_model.Load(str(_A ) ) _UpperCamelCase = vocab_file # Original fairseq vocab and spm vocab must be "aligned": # Vocab | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 # -------- | ------- | ------- | ------ | ------- | --- | --- | --- | ----- | ----- | ---- # fairseq | '<s>' | '<pad>' | '</s>' | '<unk>' | ',' | '.' | '▁' | 's' | '▁de' | '-' # spm | '<unk>' | '<s>' | '</s>' | ',' | '.' | '▁' | 's' | '▁de' | '-' | '▁a' # Mimic fairseq token-to-id alignment for the first 4 token _UpperCamelCase = {'''<s>''': 0, '''<pad>''': 1, '''</s>''': 2, '''<unk>''': 3} # The first "real" token "," has position 4 in the original fairseq vocab and position 3 in the spm vocab _UpperCamelCase = 1 _UpperCamelCase = len(self.sp_model ) _UpperCamelCase = { code: self.sp_model_size + i + self.fairseq_offset for i, code in enumerate(_A ) } _UpperCamelCase = {v: k for k, v in self.lang_code_to_id.items()} _UpperCamelCase = len(self.sp_model ) + len(self.lang_code_to_id ) + self.fairseq_offset self.fairseq_tokens_to_ids.update(self.lang_code_to_id ) _UpperCamelCase = {v: k for k, v in self.fairseq_tokens_to_ids.items()} _UpperCamelCase = src_lang if src_lang is not None else '''en_XX''' _UpperCamelCase = self.lang_code_to_id[self._src_lang] _UpperCamelCase = tgt_lang self.set_src_lang_special_tokens(self._src_lang ) @property def UpperCamelCase_ ( self : List[str] ): return len(self.sp_model ) + len(self.lang_code_to_id ) + self.fairseq_offset + 1 # Plus 1 for the mask token @property def UpperCamelCase_ ( self : Optional[int] ): return self._src_lang @src_lang.setter def UpperCamelCase_ ( self : Dict , _A : str ): _UpperCamelCase = new_src_lang self.set_src_lang_special_tokens(self._src_lang ) def __getstate__( self : List[str] ): _UpperCamelCase = self.__dict__.copy() _UpperCamelCase = None return state def __setstate__( self : List[Any] , _A : Dict ): _UpperCamelCase = d # for backward compatibility if not hasattr(self , '''sp_model_kwargs''' ): _UpperCamelCase = {} _UpperCamelCase = spm.SentencePieceProcessor(**self.sp_model_kwargs ) self.sp_model.Load(self.vocab_file ) def UpperCamelCase_ ( self : Any ): _UpperCamelCase = {self.convert_ids_to_tokens(_A ): i for i in range(self.vocab_size )} vocab.update(self.added_tokens_encoder ) return vocab def UpperCamelCase_ ( self : str , _A : str ): return self.sp_model.encode(_A , out_type=_A ) def UpperCamelCase_ ( self : Dict , _A : str ): if token in self.fairseq_tokens_to_ids: return self.fairseq_tokens_to_ids[token] _UpperCamelCase = self.sp_model.PieceToId(_A ) # Need to return unknown token if the SP model returned 0 return spm_id + self.fairseq_offset if spm_id else self.unk_token_id def UpperCamelCase_ ( self : int , _A : int ): if index in self.fairseq_ids_to_tokens: return self.fairseq_ids_to_tokens[index] return self.sp_model.IdToPiece(index - self.fairseq_offset ) def UpperCamelCase_ ( self : List[Any] , _A : str ): _UpperCamelCase = [] _UpperCamelCase = '''''' _UpperCamelCase = False for token in tokens: # make sure that special tokens are not decoded using sentencepiece model if token in self.all_special_tokens: if not prev_is_special: out_string += " " out_string += self.sp_model.decode(_A ) + token _UpperCamelCase = True _UpperCamelCase = [] else: current_sub_tokens.append(_A ) _UpperCamelCase = False out_string += self.sp_model.decode(_A ) return out_string.strip() def UpperCamelCase_ ( self : Optional[int] , _A : str , _A : Optional[str] = None ): if not os.path.isdir(_A ): logger.error(F"""Vocabulary path ({save_directory}) should be a directory""" ) return _UpperCamelCase = os.path.join( _A , (filename_prefix + '''-''' if filename_prefix else '''''') + VOCAB_FILES_NAMES['''vocab_file'''] ) if os.path.abspath(self.vocab_file ) != os.path.abspath(_A ) and os.path.isfile(self.vocab_file ): copyfile(self.vocab_file , _A ) elif not os.path.isfile(self.vocab_file ): with open(_A , '''wb''' ) as fi: _UpperCamelCase = self.sp_model.serialized_model_proto() fi.write(_A ) return (out_vocab_file,) def UpperCamelCase_ ( self : Dict , _A : List[int] , _A : Optional[List[int]] = None , _A : bool = False ): if already_has_special_tokens: return super().get_special_tokens_mask( token_ids_a=_A , token_ids_a=_A , already_has_special_tokens=_A ) _UpperCamelCase = [1] * len(self.prefix_tokens ) _UpperCamelCase = [1] * len(self.suffix_tokens ) if token_ids_a is None: return prefix_ones + ([0] * len(_A )) + suffix_ones return prefix_ones + ([0] * len(_A )) + ([0] * len(_A )) + suffix_ones def UpperCamelCase_ ( self : Tuple , _A : List[int] , _A : Optional[List[int]] = None ): if token_ids_a is None: return self.prefix_tokens + token_ids_a + self.suffix_tokens # We don't expect to process pairs, but leave the pair logic for API consistency return self.prefix_tokens + token_ids_a + token_ids_a + self.suffix_tokens def UpperCamelCase_ ( self : str , _A : Optional[int] , _A : str , _A : Optional[str] , _A : Optional[str] , **_A : List[Any] ): if src_lang is None or tgt_lang is None: raise ValueError('''Translation requires a `src_lang` and a `tgt_lang` for this model''' ) _UpperCamelCase = src_lang _UpperCamelCase = self(_A , add_special_tokens=_A , return_tensors=_A , **_A ) _UpperCamelCase = self.convert_tokens_to_ids(_A ) _UpperCamelCase = tgt_lang_id return inputs def UpperCamelCase_ ( self : int , _A : List[str] , _A : str = "en_XX" , _A : Optional[List[str]] = None , _A : str = "ro_RO" , **_A : List[str] , ): _UpperCamelCase = src_lang _UpperCamelCase = tgt_lang return super().prepare_seqaseq_batch(_A , _A , **_A ) def UpperCamelCase_ ( self : str ): return self.set_src_lang_special_tokens(self.src_lang ) def UpperCamelCase_ ( self : List[Any] ): return self.set_tgt_lang_special_tokens(self.tgt_lang ) def UpperCamelCase_ ( self : Tuple , _A : str ): _UpperCamelCase = self.lang_code_to_id[src_lang] _UpperCamelCase = [self.cur_lang_code_id] _UpperCamelCase = [self.eos_token_id] def UpperCamelCase_ ( self : List[str] , _A : str ): _UpperCamelCase = self.lang_code_to_id[tgt_lang] _UpperCamelCase = [self.cur_lang_code_id] _UpperCamelCase = [self.eos_token_id]
71
from ...configuration_utils import PretrainedConfig from ...utils import logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "EleutherAI/gpt-neox-20b": "https://huggingface.co/EleutherAI/gpt-neox-20b/resolve/main/config.json", # See all GPTNeoX models at https://huggingface.co/models?filter=gpt_neox } class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "gpt_neox" def __init__( self : Union[str, Any] , _A : Union[str, Any]=5_0432 , _A : List[Any]=6144 , _A : int=44 , _A : int=64 , _A : Optional[Any]=2_4576 , _A : Any="gelu" , _A : Tuple=0.25 , _A : Union[str, Any]=1_0000 , _A : Tuple=0.0 , _A : Any=0.0 , _A : int=0.1 , _A : List[str]=2048 , _A : Dict=0.02 , _A : Optional[Any]=1e-5 , _A : Tuple=True , _A : List[Any]=0 , _A : Optional[int]=2 , _A : Optional[int]=False , _A : List[Any]=True , _A : Any=None , **_A : Any , ): super().__init__(bos_token_id=_A , eos_token_id=_A , **_A ) _UpperCamelCase = vocab_size _UpperCamelCase = max_position_embeddings _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = rotary_pct _UpperCamelCase = rotary_emb_base _UpperCamelCase = attention_dropout _UpperCamelCase = hidden_dropout _UpperCamelCase = classifier_dropout _UpperCamelCase = initializer_range _UpperCamelCase = layer_norm_eps _UpperCamelCase = use_cache _UpperCamelCase = tie_word_embeddings _UpperCamelCase = use_parallel_residual _UpperCamelCase = rope_scaling self._rope_scaling_validation() if self.hidden_size % self.num_attention_heads != 0: raise ValueError( '''The hidden size is not divisble by the number of attention heads! Make sure to update them!''' ) def UpperCamelCase_ ( self : str ): if self.rope_scaling is None: return if not isinstance(self.rope_scaling , _A ) or len(self.rope_scaling ) != 2: raise ValueError( '''`rope_scaling` must be a dictionary with with two fields, `name` and `factor`, ''' F"""got {self.rope_scaling}""" ) _UpperCamelCase = self.rope_scaling.get('''type''' , _A ) _UpperCamelCase = self.rope_scaling.get('''factor''' , _A ) if rope_scaling_type is None or rope_scaling_type not in ["linear", "dynamic"]: raise ValueError( F"""`rope_scaling`'s name field must be one of ['linear', 'dynamic'], got {rope_scaling_type}""" ) if rope_scaling_factor is None or not isinstance(_A , _A ) or rope_scaling_factor <= 1.0: raise ValueError(F"""`rope_scaling`'s factor field must be an float > 1, got {rope_scaling_factor}""" )
71
1
import argparse import collections import json import os import re import string import sys import numpy as np _lowerCAmelCase = re.compile(r"\b(a|an|the)\b", re.UNICODE) _lowerCAmelCase = None def _snake_case ( ): _UpperCamelCase = argparse.ArgumentParser('''Official evaluation script for SQuAD version 2.0.''' ) parser.add_argument('''data_file''' , metavar='''data.json''' , help='''Input data JSON file.''' ) parser.add_argument('''pred_file''' , metavar='''pred.json''' , help='''Model predictions.''' ) parser.add_argument( '''--out-file''' , '''-o''' , metavar='''eval.json''' , help='''Write accuracy metrics to file (default is stdout).''' ) parser.add_argument( '''--na-prob-file''' , '''-n''' , metavar='''na_prob.json''' , help='''Model estimates of probability of no answer.''' ) parser.add_argument( '''--na-prob-thresh''' , '''-t''' , type=__snake_case , default=1.0 , help='''Predict "" if no-answer probability exceeds this (default = 1.0).''' , ) parser.add_argument( '''--out-image-dir''' , '''-p''' , metavar='''out_images''' , default=__snake_case , help='''Save precision-recall curves to directory.''' ) parser.add_argument('''--verbose''' , '''-v''' , action='''store_true''' ) if len(sys.argv ) == 1: parser.print_help() sys.exit(1 ) return parser.parse_args() def _snake_case ( __snake_case ): _UpperCamelCase = {} for article in dataset: for p in article["paragraphs"]: for qa in p["qas"]: _UpperCamelCase = bool(qa['''answers''']['''text'''] ) return qid_to_has_ans def _snake_case ( __snake_case ): def remove_articles(__snake_case ): return ARTICLES_REGEX.sub(''' ''' , __snake_case ) def white_space_fix(__snake_case ): return " ".join(text.split() ) def remove_punc(__snake_case ): _UpperCamelCase = set(string.punctuation ) return "".join(ch for ch in text if ch not in exclude ) def lower(__snake_case ): return text.lower() return white_space_fix(remove_articles(remove_punc(lower(__snake_case ) ) ) ) def _snake_case ( __snake_case ): if not s: return [] return normalize_answer(__snake_case ).split() def _snake_case ( __snake_case , __snake_case ): return int(normalize_answer(__snake_case ) == normalize_answer(__snake_case ) ) def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = get_tokens(__snake_case ) _UpperCamelCase = get_tokens(__snake_case ) _UpperCamelCase = collections.Counter(__snake_case ) & collections.Counter(__snake_case ) _UpperCamelCase = sum(common.values() ) if len(__snake_case ) == 0 or len(__snake_case ) == 0: # If either is no-answer, then F1 is 1 if they agree, 0 otherwise return int(gold_toks == pred_toks ) if num_same == 0: return 0 _UpperCamelCase = 1.0 * num_same / len(__snake_case ) _UpperCamelCase = 1.0 * num_same / len(__snake_case ) _UpperCamelCase = (2 * precision * recall) / (precision + recall) return fa def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = {} _UpperCamelCase = {} for article in dataset: for p in article["paragraphs"]: for qa in p["qas"]: _UpperCamelCase = qa['''id'''] _UpperCamelCase = [t for t in qa['''answers''']['''text'''] if normalize_answer(__snake_case )] if not gold_answers: # For unanswerable questions, only correct answer is empty string _UpperCamelCase = [''''''] if qid not in preds: print(f"""Missing prediction for {qid}""" ) continue _UpperCamelCase = preds[qid] # Take max over all gold answers _UpperCamelCase = max(compute_exact(__snake_case , __snake_case ) for a in gold_answers ) _UpperCamelCase = max(compute_fa(__snake_case , __snake_case ) for a in gold_answers ) return exact_scores, fa_scores def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = {} for qid, s in scores.items(): _UpperCamelCase = na_probs[qid] > na_prob_thresh if pred_na: _UpperCamelCase = float(not qid_to_has_ans[qid] ) else: _UpperCamelCase = s return new_scores def _snake_case ( __snake_case , __snake_case , __snake_case=None ): if not qid_list: _UpperCamelCase = len(__snake_case ) return collections.OrderedDict( [ ('''exact''', 100.0 * sum(exact_scores.values() ) / total), ('''f1''', 100.0 * sum(fa_scores.values() ) / total), ('''total''', total), ] ) else: _UpperCamelCase = len(__snake_case ) return collections.OrderedDict( [ ('''exact''', 100.0 * sum(exact_scores[k] for k in qid_list ) / total), ('''f1''', 100.0 * sum(fa_scores[k] for k in qid_list ) / total), ('''total''', total), ] ) def _snake_case ( __snake_case , __snake_case , __snake_case ): for k in new_eval: _UpperCamelCase = new_eval[k] def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): plt.step(__snake_case , __snake_case , color='''b''' , alpha=0.2 , where='''post''' ) plt.fill_between(__snake_case , __snake_case , step='''post''' , alpha=0.2 , color='''b''' ) plt.xlabel('''Recall''' ) plt.ylabel('''Precision''' ) plt.xlim([0.0, 1.05] ) plt.ylim([0.0, 1.05] ) plt.title(__snake_case ) plt.savefig(__snake_case ) plt.clf() def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case=None , __snake_case=None ): _UpperCamelCase = sorted(__snake_case , key=lambda __snake_case : na_probs[k] ) _UpperCamelCase = 0.0 _UpperCamelCase = 1.0 _UpperCamelCase = 0.0 _UpperCamelCase = [1.0] _UpperCamelCase = [0.0] _UpperCamelCase = 0.0 for i, qid in enumerate(__snake_case ): if qid_to_has_ans[qid]: true_pos += scores[qid] _UpperCamelCase = true_pos / float(i + 1 ) _UpperCamelCase = true_pos / float(__snake_case ) if i == len(__snake_case ) - 1 or na_probs[qid] != na_probs[qid_list[i + 1]]: # i.e., if we can put a threshold after this point avg_prec += cur_p * (cur_r - recalls[-1]) precisions.append(__snake_case ) recalls.append(__snake_case ) if out_image: plot_pr_curve(__snake_case , __snake_case , __snake_case , __snake_case ) return {"ap": 100.0 * avg_prec} def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): if out_image_dir and not os.path.exists(__snake_case ): os.makedirs(__snake_case ) _UpperCamelCase = sum(1 for v in qid_to_has_ans.values() if v ) if num_true_pos == 0: return _UpperCamelCase = make_precision_recall_eval( __snake_case , __snake_case , __snake_case , __snake_case , out_image=os.path.join(__snake_case , '''pr_exact.png''' ) , title='''Precision-Recall curve for Exact Match score''' , ) _UpperCamelCase = make_precision_recall_eval( __snake_case , __snake_case , __snake_case , __snake_case , out_image=os.path.join(__snake_case , '''pr_f1.png''' ) , title='''Precision-Recall curve for F1 score''' , ) _UpperCamelCase = {k: float(__snake_case ) for k, v in qid_to_has_ans.items()} _UpperCamelCase = make_precision_recall_eval( __snake_case , __snake_case , __snake_case , __snake_case , out_image=os.path.join(__snake_case , '''pr_oracle.png''' ) , title='''Oracle Precision-Recall curve (binary task of HasAns vs. NoAns)''' , ) merge_eval(__snake_case , __snake_case , '''pr_exact''' ) merge_eval(__snake_case , __snake_case , '''pr_f1''' ) merge_eval(__snake_case , __snake_case , '''pr_oracle''' ) def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): if not qid_list: return _UpperCamelCase = [na_probs[k] for k in qid_list] _UpperCamelCase = np.ones_like(__snake_case ) / float(len(__snake_case ) ) plt.hist(__snake_case , weights=__snake_case , bins=20 , range=(0.0, 1.0) ) plt.xlabel('''Model probability of no-answer''' ) plt.ylabel('''Proportion of dataset''' ) plt.title(f"""Histogram of no-answer probability: {name}""" ) plt.savefig(os.path.join(__snake_case , f"""na_prob_hist_{name}.png""" ) ) plt.clf() def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = sum(1 for k in qid_to_has_ans if not qid_to_has_ans[k] ) _UpperCamelCase = num_no_ans _UpperCamelCase = cur_score _UpperCamelCase = 0.0 _UpperCamelCase = sorted(__snake_case , key=lambda __snake_case : na_probs[k] ) for i, qid in enumerate(__snake_case ): if qid not in scores: continue if qid_to_has_ans[qid]: _UpperCamelCase = scores[qid] else: if preds[qid]: _UpperCamelCase = -1 else: _UpperCamelCase = 0 cur_score += diff if cur_score > best_score: _UpperCamelCase = cur_score _UpperCamelCase = na_probs[qid] return 100.0 * best_score / len(__snake_case ), best_thresh def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase , _UpperCamelCase = find_best_thresh(__snake_case , __snake_case , __snake_case , __snake_case ) _UpperCamelCase , _UpperCamelCase = find_best_thresh(__snake_case , __snake_case , __snake_case , __snake_case ) _UpperCamelCase = best_exact _UpperCamelCase = exact_thresh _UpperCamelCase = best_fa _UpperCamelCase = fa_thresh def _snake_case ( ): with open(OPTS.data_file ) as f: _UpperCamelCase = json.load(__snake_case ) _UpperCamelCase = dataset_json['''data'''] with open(OPTS.pred_file ) as f: _UpperCamelCase = json.load(__snake_case ) if OPTS.na_prob_file: with open(OPTS.na_prob_file ) as f: _UpperCamelCase = json.load(__snake_case ) else: _UpperCamelCase = {k: 0.0 for k in preds} _UpperCamelCase = make_qid_to_has_ans(__snake_case ) # maps qid to True/False _UpperCamelCase = [k for k, v in qid_to_has_ans.items() if v] _UpperCamelCase = [k for k, v in qid_to_has_ans.items() if not v] _UpperCamelCase , _UpperCamelCase = get_raw_scores(__snake_case , __snake_case ) _UpperCamelCase = apply_no_ans_threshold(__snake_case , __snake_case , __snake_case , OPTS.na_prob_thresh ) _UpperCamelCase = apply_no_ans_threshold(__snake_case , __snake_case , __snake_case , OPTS.na_prob_thresh ) _UpperCamelCase = make_eval_dict(__snake_case , __snake_case ) if has_ans_qids: _UpperCamelCase = make_eval_dict(__snake_case , __snake_case , qid_list=__snake_case ) merge_eval(__snake_case , __snake_case , '''HasAns''' ) if no_ans_qids: _UpperCamelCase = make_eval_dict(__snake_case , __snake_case , qid_list=__snake_case ) merge_eval(__snake_case , __snake_case , '''NoAns''' ) if OPTS.na_prob_file: find_all_best_thresh(__snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ) if OPTS.na_prob_file and OPTS.out_image_dir: run_precision_recall_analysis(__snake_case , __snake_case , __snake_case , __snake_case , __snake_case , OPTS.out_image_dir ) histogram_na_prob(__snake_case , __snake_case , OPTS.out_image_dir , '''hasAns''' ) histogram_na_prob(__snake_case , __snake_case , OPTS.out_image_dir , '''noAns''' ) if OPTS.out_file: with open(OPTS.out_file , '''w''' ) as f: json.dump(__snake_case , __snake_case ) else: print(json.dumps(__snake_case , indent=2 ) ) if __name__ == "__main__": _lowerCAmelCase = parse_args() if OPTS.out_image_dir: import matplotlib matplotlib.use("Agg") import matplotlib.pyplot as plt main()
71
from ..utils import DummyObject, requires_backends class lowerCAmelCase_ ( metaclass=__lowercase ): UpperCAmelCase = ["keras_nlp"] def __init__( self : Any , *_A : Dict , **_A : List[str] ): requires_backends(self , ['''keras_nlp'''] )
71
1
import argparse import os import evaluate import torch from datasets import load_dataset from torch.optim import AdamW from torch.utils.data import DataLoader from transformers import AutoModelForSequenceClassification, AutoTokenizer, get_linear_schedule_with_warmup, set_seed from accelerate import Accelerator, DistributedType ######################################################################## # This is a fully working simple example to use Accelerate, # specifically showcasing how to properly calculate the metrics on the # validation dataset when in a distributed system, and builds off the # `nlp_example.py` script. # # This example trains a Bert base model on GLUE MRPC # in any of the following settings (with the same script): # - single CPU or single GPU # - multi GPUS (using PyTorch distributed mode) # - (multi) TPUs # - fp16 (mixed-precision) or fp32 (normal precision) # # To help focus on the differences in the code, building `DataLoaders` # was refactored into its own function. # New additions from the base script can be found quickly by # looking for the # New Code # tags # # To run it in each of these various modes, follow the instructions # in the readme for examples: # https://github.com/huggingface/accelerate/tree/main/examples # ######################################################################## _lowerCAmelCase = 16 _lowerCAmelCase = 32 def _snake_case ( __snake_case , __snake_case = 16 ): _UpperCamelCase = AutoTokenizer.from_pretrained('''bert-base-cased''' ) _UpperCamelCase = load_dataset('''glue''' , '''mrpc''' ) def tokenize_function(__snake_case ): # max_length=None => use the model max length (it's actually the default) _UpperCamelCase = tokenizer(examples['''sentence1'''] , examples['''sentence2'''] , truncation=__snake_case , max_length=__snake_case ) return outputs # Apply the method we just defined to all the examples in all the splits of the dataset # starting with the main process first: with accelerator.main_process_first(): _UpperCamelCase = datasets.map( __snake_case , batched=__snake_case , remove_columns=['''idx''', '''sentence1''', '''sentence2'''] , ) # We also rename the 'label' column to 'labels' which is the expected name for labels by the models of the # transformers library _UpperCamelCase = tokenized_datasets.rename_column('''label''' , '''labels''' ) def collate_fn(__snake_case ): # On TPU it's best to pad everything to the same length or training will be very slow. _UpperCamelCase = 128 if accelerator.distributed_type == DistributedType.TPU else None # When using mixed precision we want round multiples of 8/16 if accelerator.mixed_precision == "fp8": _UpperCamelCase = 16 elif accelerator.mixed_precision != "no": _UpperCamelCase = 8 else: _UpperCamelCase = None return tokenizer.pad( __snake_case , padding='''longest''' , max_length=__snake_case , pad_to_multiple_of=__snake_case , return_tensors='''pt''' , ) # Instantiate dataloaders. _UpperCamelCase = DataLoader( tokenized_datasets['''train'''] , shuffle=__snake_case , collate_fn=__snake_case , batch_size=__snake_case ) _UpperCamelCase = DataLoader( tokenized_datasets['''validation'''] , shuffle=__snake_case , collate_fn=__snake_case , batch_size=__snake_case ) return train_dataloader, eval_dataloader # For testing only if os.environ.get("TESTING_MOCKED_DATALOADERS", None) == "1": from accelerate.test_utils.training import mocked_dataloaders _lowerCAmelCase = mocked_dataloaders # noqa: F811 def _snake_case ( __snake_case , __snake_case ): # For testing only if os.environ.get('''TESTING_MOCKED_DATALOADERS''' , __snake_case ) == "1": _UpperCamelCase = 2 # Initialize accelerator _UpperCamelCase = Accelerator(cpu=args.cpu , mixed_precision=args.mixed_precision ) # Sample hyper-parameters for learning rate, batch size, seed and a few other HPs _UpperCamelCase = config['''lr'''] _UpperCamelCase = int(config['''num_epochs'''] ) _UpperCamelCase = int(config['''seed'''] ) _UpperCamelCase = int(config['''batch_size'''] ) _UpperCamelCase = evaluate.load('''glue''' , '''mrpc''' ) # If the batch size is too big we use gradient accumulation _UpperCamelCase = 1 if batch_size > MAX_GPU_BATCH_SIZE and accelerator.distributed_type != DistributedType.TPU: _UpperCamelCase = batch_size // MAX_GPU_BATCH_SIZE _UpperCamelCase = MAX_GPU_BATCH_SIZE set_seed(__snake_case ) _UpperCamelCase , _UpperCamelCase = get_dataloaders(__snake_case , __snake_case ) # Instantiate the model (we build the model here so that the seed also control new weights initialization) _UpperCamelCase = AutoModelForSequenceClassification.from_pretrained('''bert-base-cased''' , return_dict=__snake_case ) # We could avoid this line since the accelerator is set with `device_placement=True` (default value). # Note that if you are placing tensors on devices manually, this line absolutely needs to be before the optimizer # creation otherwise training will not work on TPU (`accelerate` will kindly throw an error to make us aware of that). _UpperCamelCase = model.to(accelerator.device ) # Instantiate optimizer _UpperCamelCase = AdamW(params=model.parameters() , lr=__snake_case ) # Instantiate scheduler _UpperCamelCase = get_linear_schedule_with_warmup( optimizer=__snake_case , num_warmup_steps=100 , num_training_steps=(len(__snake_case ) * num_epochs) // gradient_accumulation_steps , ) # Prepare everything # There is no specific order to remember, we just need to unpack the objects in the same order we gave them to the # prepare method. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = accelerator.prepare( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ) # Now we train the model for epoch in range(__snake_case ): model.train() for step, batch in enumerate(__snake_case ): # We could avoid this line since we set the accelerator with `device_placement=True`. batch.to(accelerator.device ) _UpperCamelCase = model(**__snake_case ) _UpperCamelCase = outputs.loss _UpperCamelCase = loss / gradient_accumulation_steps accelerator.backward(__snake_case ) if step % gradient_accumulation_steps == 0: optimizer.step() lr_scheduler.step() optimizer.zero_grad() model.eval() _UpperCamelCase = 0 for step, batch in enumerate(__snake_case ): # We could avoid this line since we set the accelerator with `device_placement=True`. batch.to(accelerator.device ) with torch.no_grad(): _UpperCamelCase = model(**__snake_case ) _UpperCamelCase = outputs.logits.argmax(dim=-1 ) _UpperCamelCase , _UpperCamelCase = accelerator.gather((predictions, batch['''labels''']) ) # New Code # # First we check if it's a distributed system if accelerator.use_distributed: # Then see if we're on the last batch of our eval dataloader if step == len(__snake_case ) - 1: # Last batch needs to be truncated on distributed systems as it contains additional samples _UpperCamelCase = predictions[: len(eval_dataloader.dataset ) - samples_seen] _UpperCamelCase = references[: len(eval_dataloader.dataset ) - samples_seen] else: # Otherwise we add the number of samples seen samples_seen += references.shape[0] # All of this can be avoided if you use `Accelerator.gather_for_metrics` instead of `Accelerator.gather`: # accelerator.gather_for_metrics((predictions, batch["labels"])) metric.add_batch( predictions=__snake_case , references=__snake_case , ) _UpperCamelCase = metric.compute() # Use accelerator.print to print only on the main process. accelerator.print(f"""epoch {epoch}:""" , __snake_case ) def _snake_case ( ): _UpperCamelCase = argparse.ArgumentParser(description='''Simple example of training script.''' ) parser.add_argument( '''--mixed_precision''' , type=__snake_case , default=__snake_case , choices=['''no''', '''fp16''', '''bf16''', '''fp8'''] , help='''Whether to use mixed precision. Choose''' '''between fp16 and bf16 (bfloat16). Bf16 requires PyTorch >= 1.10.''' '''and an Nvidia Ampere GPU.''' , ) parser.add_argument('''--cpu''' , action='''store_true''' , help='''If passed, will train on the CPU.''' ) _UpperCamelCase = parser.parse_args() _UpperCamelCase = {'''lr''': 2E-5, '''num_epochs''': 3, '''seed''': 42, '''batch_size''': 16} training_function(__snake_case , __snake_case ) if __name__ == "__main__": main()
71
from typing import Optional, Tuple, Union import tensorflow as tf from ...activations_tf import ACTaFN from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward from ...modeling_tf_outputs import ( TFBaseModelOutputWithNoAttention, TFBaseModelOutputWithPoolingAndNoAttention, TFSequenceClassifierOutput, ) from ...modeling_tf_utils import TFPreTrainedModel, TFSequenceClassificationLoss, keras_serializable, unpack_inputs from ...tf_utils import shape_list from ...utils import logging from .configuration_regnet import RegNetConfig _lowerCAmelCase = logging.get_logger(__name__) # General docstring _lowerCAmelCase = "RegNetConfig" # Base docstring _lowerCAmelCase = "facebook/regnet-y-040" _lowerCAmelCase = [1, 1_088, 7, 7] # Image classification docstring _lowerCAmelCase = "facebook/regnet-y-040" _lowerCAmelCase = "tabby, tabby cat" _lowerCAmelCase = [ "facebook/regnet-y-040", # See all regnet models at https://huggingface.co/models?filter=regnet ] class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : str , _A : int , _A : int = 3 , _A : int = 1 , _A : int = 1 , _A : Optional[str] = "relu" , **_A : Any , ): super().__init__(**_A ) # The padding and conv has been verified in # https://colab.research.google.com/gist/sayakpaul/854bc10eeaf21c9ee2119e0b9f3841a7/scratchpad.ipynb _UpperCamelCase = tf.keras.layers.ZeroPaddingaD(padding=kernel_size // 2 ) _UpperCamelCase = tf.keras.layers.ConvaD( filters=_A , kernel_size=_A , strides=_A , padding='''VALID''' , groups=_A , use_bias=_A , name='''convolution''' , ) _UpperCamelCase = tf.keras.layers.BatchNormalization(epsilon=1e-5 , momentum=0.9 , name='''normalization''' ) _UpperCamelCase = ACTaFN[activation] if activation is not None else tf.identity def UpperCamelCase_ ( self : Any , _A : Any ): _UpperCamelCase = self.convolution(self.padding(_A ) ) _UpperCamelCase = self.normalization(_A ) _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Optional[Any] , _A : RegNetConfig , **_A : Any ): super().__init__(**_A ) _UpperCamelCase = config.num_channels _UpperCamelCase = TFRegNetConvLayer( out_channels=config.embedding_size , kernel_size=3 , stride=2 , activation=config.hidden_act , name='''embedder''' , ) def UpperCamelCase_ ( self : List[str] , _A : Optional[int] ): _UpperCamelCase = shape_list(_A )[1] if tf.executing_eagerly() and num_channels != self.num_channels: raise ValueError( '''Make sure that the channel dimension of the pixel values match with the one set in the configuration.''' ) # When running on CPU, `tf.keras.layers.Conv2D` doesn't support `NCHW` format. # So change the input format from `NCHW` to `NHWC`. # shape = (batch_size, in_height, in_width, in_channels=num_channels) _UpperCamelCase = tf.transpose(_A , perm=(0, 2, 3, 1) ) _UpperCamelCase = self.embedder(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : str , _A : int , _A : int = 2 , **_A : Optional[Any] ): super().__init__(**_A ) _UpperCamelCase = tf.keras.layers.ConvaD( filters=_A , kernel_size=1 , strides=_A , use_bias=_A , name='''convolution''' ) _UpperCamelCase = tf.keras.layers.BatchNormalization(epsilon=1e-5 , momentum=0.9 , name='''normalization''' ) def UpperCamelCase_ ( self : str , _A : tf.Tensor , _A : bool = False ): return self.normalization(self.convolution(_A ) , training=_A ) class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Dict , _A : int , _A : int , **_A : Dict ): super().__init__(**_A ) _UpperCamelCase = tf.keras.layers.GlobalAveragePoolingaD(keepdims=_A , name='''pooler''' ) _UpperCamelCase = [ tf.keras.layers.ConvaD(filters=_A , kernel_size=1 , activation='''relu''' , name='''attention.0''' ), tf.keras.layers.ConvaD(filters=_A , kernel_size=1 , activation='''sigmoid''' , name='''attention.2''' ), ] def UpperCamelCase_ ( self : List[str] , _A : List[Any] ): # [batch_size, h, w, num_channels] -> [batch_size, 1, 1, num_channels] _UpperCamelCase = self.pooler(_A ) for layer_module in self.attention: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = hidden_state * pooled return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , _A : int , _A : int , _A : int = 1 , **_A : str ): super().__init__(**_A ) _UpperCamelCase = in_channels != out_channels or stride != 1 _UpperCamelCase = max(1 , out_channels // config.groups_width ) _UpperCamelCase = ( TFRegNetShortCut(_A , stride=_A , name='''shortcut''' ) if should_apply_shortcut else tf.keras.layers.Activation('''linear''' , name='''shortcut''' ) ) # `self.layers` instead of `self.layer` because that is a reserved argument. _UpperCamelCase = [ TFRegNetConvLayer(_A , kernel_size=1 , activation=config.hidden_act , name='''layer.0''' ), TFRegNetConvLayer( _A , stride=_A , groups=_A , activation=config.hidden_act , name='''layer.1''' ), TFRegNetConvLayer(_A , kernel_size=1 , activation=_A , name='''layer.2''' ), ] _UpperCamelCase = ACTaFN[config.hidden_act] def UpperCamelCase_ ( self : Dict , _A : Tuple ): _UpperCamelCase = hidden_state for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = self.shortcut(_A ) hidden_state += residual _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , _A : int , _A : int , _A : int = 1 , **_A : int ): super().__init__(**_A ) _UpperCamelCase = in_channels != out_channels or stride != 1 _UpperCamelCase = max(1 , out_channels // config.groups_width ) _UpperCamelCase = ( TFRegNetShortCut(_A , stride=_A , name='''shortcut''' ) if should_apply_shortcut else tf.keras.layers.Activation('''linear''' , name='''shortcut''' ) ) _UpperCamelCase = [ TFRegNetConvLayer(_A , kernel_size=1 , activation=config.hidden_act , name='''layer.0''' ), TFRegNetConvLayer( _A , stride=_A , groups=_A , activation=config.hidden_act , name='''layer.1''' ), TFRegNetSELayer(_A , reduced_channels=int(round(in_channels / 4 ) ) , name='''layer.2''' ), TFRegNetConvLayer(_A , kernel_size=1 , activation=_A , name='''layer.3''' ), ] _UpperCamelCase = ACTaFN[config.hidden_act] def UpperCamelCase_ ( self : Tuple , _A : List[Any] ): _UpperCamelCase = hidden_state for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = self.shortcut(_A ) hidden_state += residual _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Tuple , _A : RegNetConfig , _A : int , _A : int , _A : int = 2 , _A : int = 2 , **_A : Union[str, Any] ): super().__init__(**_A ) _UpperCamelCase = TFRegNetXLayer if config.layer_type == '''x''' else TFRegNetYLayer _UpperCamelCase = [ # downsampling is done in the first layer with stride of 2 layer(_A , _A , _A , stride=_A , name='''layers.0''' ), *[layer(_A , _A , _A , name=F"""layers.{i+1}""" ) for i in range(depth - 1 )], ] def UpperCamelCase_ ( self : Union[str, Any] , _A : Optional[int] ): for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , **_A : List[str] ): super().__init__(**_A ) _UpperCamelCase = [] # based on `downsample_in_first_stage`, the first layer of the first stage may or may not downsample the input self.stages.append( TFRegNetStage( _A , config.embedding_size , config.hidden_sizes[0] , stride=2 if config.downsample_in_first_stage else 1 , depth=config.depths[0] , name='''stages.0''' , ) ) _UpperCamelCase = zip(config.hidden_sizes , config.hidden_sizes[1:] ) for i, ((in_channels, out_channels), depth) in enumerate(zip(_A , config.depths[1:] ) ): self.stages.append(TFRegNetStage(_A , _A , _A , depth=_A , name=F"""stages.{i+1}""" ) ) def UpperCamelCase_ ( self : Optional[int] , _A : tf.Tensor , _A : bool = False , _A : bool = True ): _UpperCamelCase = () if output_hidden_states else None for stage_module in self.stages: if output_hidden_states: _UpperCamelCase = hidden_states + (hidden_state,) _UpperCamelCase = stage_module(_A ) if output_hidden_states: _UpperCamelCase = hidden_states + (hidden_state,) if not return_dict: return tuple(v for v in [hidden_state, hidden_states] if v is not None ) return TFBaseModelOutputWithNoAttention(last_hidden_state=_A , hidden_states=_A ) @keras_serializable class lowerCAmelCase_ ( tf.keras.layers.Layer ): UpperCAmelCase = RegNetConfig def __init__( self : int , _A : Tuple , **_A : int ): super().__init__(**_A ) _UpperCamelCase = config _UpperCamelCase = TFRegNetEmbeddings(_A , name='''embedder''' ) _UpperCamelCase = TFRegNetEncoder(_A , name='''encoder''' ) _UpperCamelCase = tf.keras.layers.GlobalAveragePoolingaD(keepdims=_A , name='''pooler''' ) @unpack_inputs def UpperCamelCase_ ( self : Optional[int] , _A : tf.Tensor , _A : Optional[bool] = None , _A : Optional[bool] = None , _A : bool = False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.embedder(_A , training=_A ) _UpperCamelCase = self.encoder( _A , output_hidden_states=_A , return_dict=_A , training=_A ) _UpperCamelCase = encoder_outputs[0] _UpperCamelCase = self.pooler(_A ) # Change to NCHW output format have uniformity in the modules _UpperCamelCase = tf.transpose(_A , perm=(0, 3, 1, 2) ) _UpperCamelCase = tf.transpose(_A , perm=(0, 3, 1, 2) ) # Change the other hidden state outputs to NCHW as well if output_hidden_states: _UpperCamelCase = tuple([tf.transpose(_A , perm=(0, 3, 1, 2) ) for h in encoder_outputs[1]] ) if not return_dict: return (last_hidden_state, pooled_output) + encoder_outputs[1:] return TFBaseModelOutputWithPoolingAndNoAttention( last_hidden_state=_A , pooler_output=_A , hidden_states=hidden_states if output_hidden_states else encoder_outputs.hidden_states , ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = RegNetConfig UpperCAmelCase = "regnet" UpperCAmelCase = "pixel_values" @property def UpperCamelCase_ ( self : Tuple ): return {"pixel_values": tf.TensorSpec(shape=(None, self.config.num_channels, 224, 224) , dtype=tf.floataa )} _lowerCAmelCase = r"\n Parameters:\n This model is a Tensorflow\n [tf.keras.layers.Layer](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Layer) sub-class. Use it as a\n regular Tensorflow Module and refer to the Tensorflow documentation for all matter related to general usage and\n behavior.\n config ([`RegNetConfig`]): Model configuration class with all the parameters of the model.\n Initializing with a config file does not load the weights associated with the model, only the\n configuration. Check out the [`~TFPreTrainedModel.from_pretrained`] method to load the model weights.\n" _lowerCAmelCase = r"\n Args:\n pixel_values (`tf.Tensor` of shape `(batch_size, num_channels, height, width)`):\n Pixel values. Pixel values can be obtained using [`AutoImageProcessor`]. See\n [`ConveNextImageProcessor.__call__`] for details.\n output_hidden_states (`bool`, *optional*):\n Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for\n more detail.\n return_dict (`bool`, *optional*):\n Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.\n" @add_start_docstrings( "The bare RegNet model outputting raw features without any specific head on top.", __lowercase, ) class lowerCAmelCase_ ( __lowercase ): def __init__( self : List[Any] , _A : RegNetConfig , *_A : Optional[int] , **_A : Tuple ): super().__init__(_A , *_A , **_A ) _UpperCamelCase = TFRegNetMainLayer(_A , name='''regnet''' ) @unpack_inputs @add_start_docstrings_to_model_forward(_A ) @add_code_sample_docstrings( checkpoint=_CHECKPOINT_FOR_DOC , output_type=_A , config_class=_CONFIG_FOR_DOC , modality='''vision''' , expected_output=_EXPECTED_OUTPUT_SHAPE , ) def UpperCamelCase_ ( self : Any , _A : tf.Tensor , _A : Optional[bool] = None , _A : Optional[bool] = None , _A : Optional[int]=False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.regnet( pixel_values=_A , output_hidden_states=_A , return_dict=_A , training=_A , ) if not return_dict: return (outputs[0],) + outputs[1:] return TFBaseModelOutputWithPoolingAndNoAttention( last_hidden_state=outputs.last_hidden_state , pooler_output=outputs.pooler_output , hidden_states=outputs.hidden_states , ) @add_start_docstrings( "\n RegNet Model with an image classification head on top (a linear layer on top of the pooled features), e.g. for\n ImageNet.\n ", __lowercase, ) class lowerCAmelCase_ ( __lowercase, __lowercase ): def __init__( self : List[Any] , _A : RegNetConfig , *_A : Any , **_A : int ): super().__init__(_A , *_A , **_A ) _UpperCamelCase = config.num_labels _UpperCamelCase = TFRegNetMainLayer(_A , name='''regnet''' ) # classification head _UpperCamelCase = [ tf.keras.layers.Flatten(), tf.keras.layers.Dense(config.num_labels , name='''classifier.1''' ) if config.num_labels > 0 else tf.identity, ] @unpack_inputs @add_start_docstrings_to_model_forward(_A ) @add_code_sample_docstrings( checkpoint=_IMAGE_CLASS_CHECKPOINT , output_type=_A , config_class=_CONFIG_FOR_DOC , expected_output=_IMAGE_CLASS_EXPECTED_OUTPUT , ) def UpperCamelCase_ ( self : str , _A : tf.Tensor = None , _A : tf.Tensor = None , _A : bool = None , _A : bool = None , _A : Any=False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.regnet( _A , output_hidden_states=_A , return_dict=_A , training=_A ) _UpperCamelCase = outputs.pooler_output if return_dict else outputs[1] _UpperCamelCase = self.classifier[0](_A ) _UpperCamelCase = self.classifier[1](_A ) _UpperCamelCase = None if labels is None else self.hf_compute_loss(labels=_A , logits=_A ) if not return_dict: _UpperCamelCase = (logits,) + outputs[2:] return ((loss,) + output) if loss is not None else output return TFSequenceClassifierOutput(loss=_A , logits=_A , hidden_states=outputs.hidden_states )
71
1
from collections import namedtuple import requests from lxml import html # type: ignore _lowerCAmelCase = namedtuple("covid_data", "cases deaths recovered") def _snake_case ( __snake_case = "https://www.worldometers.info/coronavirus/" ): _UpperCamelCase = '''//div[@class = "maincounter-number"]/span/text()''' return covid_data(*html.fromstring(requests.get(__snake_case ).content ).xpath(__snake_case ) ) _lowerCAmelCase = "Total COVID-19 cases in the world: {}\nTotal deaths due to COVID-19 in the world: {}\nTotal COVID-19 patients recovered in the world: {}" print(fmt.format(*covid_stats()))
71
from sklearn.metrics import mean_squared_error import datasets _lowerCAmelCase = "\\n@article{scikit-learn,\n title={Scikit-learn: Machine Learning in {P}ython},\n author={Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V.\n and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P.\n and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and\n Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E.},\n journal={Journal of Machine Learning Research},\n volume={12},\n pages={2825--2830},\n year={2011}\n}\n" _lowerCAmelCase = "\\nMean Squared Error(MSE) is the average of the square of difference between the predicted\nand actual values.\n" _lowerCAmelCase = "\nArgs:\n predictions: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Estimated target values.\n references: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Ground truth (correct) target values.\n sample_weight: array-like of shape (n_samples,), default=None\n Sample weights.\n multioutput: {\"raw_values\", \"uniform_average\"} or array-like of shape (n_outputs,), default=\"uniform_average\"\n Defines aggregating of multiple output values. Array-like value defines weights used to average errors.\n\n \"raw_values\" : Returns a full set of errors in case of multioutput input.\n\n \"uniform_average\" : Errors of all outputs are averaged with uniform weight.\n\n squared : bool, default=True\n If True returns MSE value, if False returns RMSE (Root Mean Squared Error) value.\n\nReturns:\n mse : mean squared error.\nExamples:\n\n >>> mse_metric = datasets.load_metric(\"mse\")\n >>> predictions = [2.5, 0.0, 2, 8]\n >>> references = [3, -0.5, 2, 7]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.375}\n >>> rmse_result = mse_metric.compute(predictions=predictions, references=references, squared=False)\n >>> print(rmse_result)\n {'mse': 0.6123724356957945}\n\n If you're using multi-dimensional lists, then set the config as follows :\n\n >>> mse_metric = datasets.load_metric(\"mse\", \"multilist\")\n >>> predictions = [[0.5, 1], [-1, 1], [7, -6]]\n >>> references = [[0, 2], [-1, 2], [8, -5]]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.7083333333333334}\n >>> results = mse_metric.compute(predictions=predictions, references=references, multioutput='raw_values')\n >>> print(results) # doctest: +NORMALIZE_WHITESPACE\n {'mse': array([0.41666667, 1. ])}\n" @datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION ) class lowerCAmelCase_ ( datasets.Metric ): def UpperCamelCase_ ( self : Optional[int] ): return datasets.MetricInfo( description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features(self._get_feature_types() ) , reference_urls=[ '''https://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html''' ] , ) def UpperCamelCase_ ( self : Dict ): if self.config_name == "multilist": return { "predictions": datasets.Sequence(datasets.Value('''float''' ) ), "references": datasets.Sequence(datasets.Value('''float''' ) ), } else: return { "predictions": datasets.Value('''float''' ), "references": datasets.Value('''float''' ), } def UpperCamelCase_ ( self : Any , _A : List[Any] , _A : List[str] , _A : Dict=None , _A : List[str]="uniform_average" , _A : int=True ): _UpperCamelCase = mean_squared_error( _A , _A , sample_weight=_A , multioutput=_A , squared=_A ) return {"mse": mse}
71
1
import copy import os from typing import Union from ...configuration_utils import PretrainedConfig from ...utils import logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "google/pix2struct-textcaps-base": ( "https://huggingface.co/google/pix2struct-textcaps-base/resolve/main/config.json" ), } class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "pix2struct_text_model" UpperCAmelCase = ["past_key_values"] UpperCAmelCase = { "hidden_size": "hidden_size", "num_attention_heads": "num_heads", "num_hidden_layers": "num_layers", } def __init__( self : List[Any] , _A : Union[str, Any]=5_0244 , _A : Dict=768 , _A : str=64 , _A : Optional[Any]=2048 , _A : Optional[Any]=12 , _A : str=12 , _A : List[str]=32 , _A : int=128 , _A : Tuple=0.1 , _A : Union[str, Any]=1e-6 , _A : str=1.0 , _A : Optional[int]="gelu_new" , _A : Tuple=0 , _A : List[str]=False , _A : Optional[Any]=0 , _A : Tuple=1 , _A : Tuple=False , _A : List[Any]=True , **_A : List[Any] , ): _UpperCamelCase = vocab_size _UpperCamelCase = hidden_size _UpperCamelCase = d_kv _UpperCamelCase = d_ff _UpperCamelCase = num_layers _UpperCamelCase = num_heads _UpperCamelCase = relative_attention_num_buckets _UpperCamelCase = relative_attention_max_distance _UpperCamelCase = dropout_rate _UpperCamelCase = layer_norm_epsilon _UpperCamelCase = initializer_factor _UpperCamelCase = use_cache _UpperCamelCase = eos_token_id _UpperCamelCase = decoder_start_token_id # for backwards compatibility _UpperCamelCase = dense_act_fn super().__init__( pad_token_id=_A , eos_token_id=_A , decoder_start_token_id=_A , tie_word_embeddings=_A , is_decoder=_A , **_A , ) @classmethod def UpperCamelCase_ ( cls : Dict , _A : Union[str, os.PathLike] , **_A : List[Any] ): cls._set_token_in_kwargs(_A ) _UpperCamelCase , _UpperCamelCase = cls.get_config_dict(_A , **_A ) # get the text config dict if we are loading from Pix2StructConfig if config_dict.get('''model_type''' ) == "pix2struct": _UpperCamelCase = config_dict['''text_config'''] if "model_type" in config_dict and hasattr(cls , '''model_type''' ) and config_dict["model_type"] != cls.model_type: logger.warning( F"""You are using a model of type {config_dict["model_type"]} to instantiate a model of type """ F"""{cls.model_type}. This is not supported for all configurations of models and can yield errors.""" ) return cls.from_dict(_A , **_A ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "pix2struct_vision_model" def __init__( self : Optional[int] , _A : str=768 , _A : str=768 , _A : Optional[Any]=2048 , _A : Optional[Any]=64 , _A : Any=12 , _A : Tuple=12 , _A : Dict="gelu_new" , _A : Any=1e-6 , _A : str=0.0 , _A : Any=0.0 , _A : Optional[Any]=1e-10 , _A : Any=1.0 , _A : int=4096 , _A : int=32 , _A : List[Any]=128 , **_A : Optional[int] , ): super().__init__(**_A ) _UpperCamelCase = hidden_size _UpperCamelCase = patch_embed_hidden_size _UpperCamelCase = d_ff _UpperCamelCase = dropout_rate _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = initializer_range _UpperCamelCase = initializer_factor _UpperCamelCase = attention_dropout _UpperCamelCase = layer_norm_eps _UpperCamelCase = dense_act_fn _UpperCamelCase = seq_len _UpperCamelCase = relative_attention_num_buckets _UpperCamelCase = relative_attention_max_distance _UpperCamelCase = d_kv @classmethod def UpperCamelCase_ ( cls : str , _A : Union[str, os.PathLike] , **_A : Any ): cls._set_token_in_kwargs(_A ) _UpperCamelCase , _UpperCamelCase = cls.get_config_dict(_A , **_A ) # get the vision config dict if we are loading from Pix2StructConfig if config_dict.get('''model_type''' ) == "pix2struct": _UpperCamelCase = config_dict['''vision_config'''] if "model_type" in config_dict and hasattr(cls , '''model_type''' ) and config_dict["model_type"] != cls.model_type: logger.warning( F"""You are using a model of type {config_dict["model_type"]} to instantiate a model of type """ F"""{cls.model_type}. This is not supported for all configurations of models and can yield errors.""" ) return cls.from_dict(_A , **_A ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "pix2struct" UpperCAmelCase = True def __init__( self : Union[str, Any] , _A : Tuple=None , _A : Any=None , _A : Dict=1.0 , _A : Optional[int]=0.02 , _A : List[Any]=False , _A : List[Any]=False , _A : Any=True , **_A : str , ): super().__init__(tie_word_embeddings=_A , is_encoder_decoder=_A , **_A ) if text_config is None: _UpperCamelCase = {} logger.info('''text_config is None. Initializing the Pix2StructTextConfig with default values.''' ) if vision_config is None: _UpperCamelCase = {} logger.info('''vision_config is None. Initializing the Pix2StructVisionConfig with default values.''' ) _UpperCamelCase = PixaStructTextConfig(**_A ) _UpperCamelCase = PixaStructVisionConfig(**_A ) _UpperCamelCase = self.text_config.decoder_start_token_id _UpperCamelCase = self.text_config.pad_token_id _UpperCamelCase = self.text_config.eos_token_id _UpperCamelCase = initializer_factor _UpperCamelCase = initializer_range _UpperCamelCase = self.initializer_range _UpperCamelCase = self.initializer_range _UpperCamelCase = is_vqa @classmethod def UpperCamelCase_ ( cls : Any , _A : PixaStructTextConfig , _A : PixaStructVisionConfig , **_A : Union[str, Any] ): return cls(text_config=text_config.to_dict() , vision_config=vision_config.to_dict() , **_A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = copy.deepcopy(self.__dict__ ) _UpperCamelCase = self.text_config.to_dict() _UpperCamelCase = self.vision_config.to_dict() _UpperCamelCase = self.__class__.model_type return output
71
import os import re import shutil import sys import tempfile import unittest import black _lowerCAmelCase = os.path.abspath(os.path.dirname(os.path.dirname(os.path.dirname(__file__)))) sys.path.append(os.path.join(git_repo_path, "utils")) import check_copies # noqa: E402 # This is the reference code that will be used in the tests. # If DDPMSchedulerOutput is changed in scheduling_ddpm.py, this code needs to be manually updated. _lowerCAmelCase = " \"\"\"\n Output class for the scheduler's step function output.\n\n Args:\n prev_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n Computed sample (x_{t-1}) of previous timestep. `prev_sample` should be used as next model input in the\n denoising loop.\n pred_original_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n The predicted denoised sample (x_{0}) based on the model output from the current timestep.\n `pred_original_sample` can be used to preview progress or for guidance.\n \"\"\"\n\n prev_sample: torch.FloatTensor\n pred_original_sample: Optional[torch.FloatTensor] = None\n" class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = tempfile.mkdtemp() os.makedirs(os.path.join(self.diffusers_dir , '''schedulers/''' ) ) _UpperCamelCase = self.diffusers_dir shutil.copy( os.path.join(_A , '''src/diffusers/schedulers/scheduling_ddpm.py''' ) , os.path.join(self.diffusers_dir , '''schedulers/scheduling_ddpm.py''' ) , ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = '''src/diffusers''' shutil.rmtree(self.diffusers_dir ) def UpperCamelCase_ ( self : Union[str, Any] , _A : Tuple , _A : Optional[Any] , _A : Dict , _A : List[str]=None ): _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + class_code if overwrite_result is not None: _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + overwrite_result _UpperCamelCase = black.Mode(target_versions={black.TargetVersion.PYaa} , line_length=119 ) _UpperCamelCase = black.format_str(_A , mode=_A ) _UpperCamelCase = os.path.join(self.diffusers_dir , '''new_code.py''' ) with open(_A , '''w''' , newline='''\n''' ) as f: f.write(_A ) if overwrite_result is None: self.assertTrue(len(check_copies.is_copy_consistent(_A ) ) == 0 ) else: check_copies.is_copy_consistent(f.name , overwrite=_A ) with open(_A , '''r''' ) as f: self.assertTrue(f.read() , _A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = check_copies.find_code_in_diffusers('''schedulers.scheduling_ddpm.DDPMSchedulerOutput''' ) self.assertEqual(_A , _A ) def UpperCamelCase_ ( self : Optional[Any] ): # Base copy consistency self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , REFERENCE_CODE + '''\n''' , ) # With no empty line at the end self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , _A , ) # Copy consistency with rename self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , re.sub('''DDPM''' , '''Test''' , _A ) , ) # Copy consistency with a really long name _UpperCamelCase = '''TestClassWithAReallyLongNameBecauseSomePeopleLikeThatForSomeReason''' self.check_copy_consistency( F"""# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->{long_class_name}""" , F"""{long_class_name}SchedulerOutput""" , re.sub('''Bert''' , _A , _A ) , ) # Copy consistency with overwrite self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , _A , overwrite_result=re.sub('''DDPM''' , '''Test''' , _A ) , )
71
1
import warnings from ...utils import logging from .image_processing_glpn import GLPNImageProcessor _lowerCAmelCase = logging.get_logger(__name__) class lowerCAmelCase_ ( __lowercase ): def __init__( self : List[Any] , *_A : Tuple , **_A : Any ): warnings.warn( '''The class GLPNFeatureExtractor is deprecated and will be removed in version 5 of Transformers. Please''' ''' use GLPNImageProcessor instead.''' , _A , ) super().__init__(*_A , **_A )
71
from __future__ import annotations import math class lowerCAmelCase_ : def __init__( self : int , _A : int ): _UpperCamelCase = size # approximate the overall size of segment tree with given value _UpperCamelCase = [0 for i in range(0 , 4 * size )] # create array to store lazy update _UpperCamelCase = [0 for i in range(0 , 4 * size )] _UpperCamelCase = [0 for i in range(0 , 4 * size )] # flag for lazy update def UpperCamelCase_ ( self : str , _A : int ): return idx * 2 def UpperCamelCase_ ( self : Any , _A : int ): return idx * 2 + 1 def UpperCamelCase_ ( self : Union[str, Any] , _A : int , _A : int , _A : int , _A : list[int] ): if left_element == right_element: _UpperCamelCase = a[left_element - 1] else: _UpperCamelCase = (left_element + right_element) // 2 self.build(self.left(_A ) , _A , _A , _A ) self.build(self.right(_A ) , mid + 1 , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) def UpperCamelCase_ ( self : Tuple , _A : int , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return True if left_element >= a and right_element <= b: _UpperCamelCase = val if left_element != right_element: _UpperCamelCase = val _UpperCamelCase = val _UpperCamelCase = True _UpperCamelCase = True return True _UpperCamelCase = (left_element + right_element) // 2 self.update(self.left(_A ) , _A , _A , _A , _A , _A ) self.update(self.right(_A ) , mid + 1 , _A , _A , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) return True def UpperCamelCase_ ( self : Any , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return -math.inf if left_element >= a and right_element <= b: return self.segment_tree[idx] _UpperCamelCase = (left_element + right_element) // 2 _UpperCamelCase = self.query(self.left(_A ) , _A , _A , _A , _A ) _UpperCamelCase = self.query(self.right(_A ) , mid + 1 , _A , _A , _A ) return max(_A , _A ) def __str__( self : Tuple ): return str([self.query(1 , 1 , self.size , _A , _A ) for i in range(1 , self.size + 1 )] ) if __name__ == "__main__": _lowerCAmelCase = [1, 2, -4, 7, 3, -5, 6, 11, -20, 9, 14, 15, 5, 2, -8] _lowerCAmelCase = 15 _lowerCAmelCase = SegmentTree(size) segt.build(1, 1, size, A) print(segt.query(1, 1, size, 4, 6)) print(segt.query(1, 1, size, 7, 11)) print(segt.query(1, 1, size, 7, 12)) segt.update(1, 1, size, 1, 3, 111) print(segt.query(1, 1, size, 1, 15)) segt.update(1, 1, size, 7, 8, 235) print(segt)
71
1
def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): # Return True if there is node that has not iterated. _UpperCamelCase = [False] * len(__snake_case ) _UpperCamelCase = [] queue.append(__snake_case ) _UpperCamelCase = True while queue: _UpperCamelCase = queue.pop(0 ) for ind in range(len(graph[u] ) ): if visited[ind] is False and graph[u][ind] > 0: queue.append(__snake_case ) _UpperCamelCase = True _UpperCamelCase = u return visited[t] def _snake_case ( __snake_case , __snake_case , __snake_case ): # This array is filled by BFS and to store path _UpperCamelCase = [-1] * (len(__snake_case )) _UpperCamelCase = 0 while bfs(__snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = float('''Inf''' ) _UpperCamelCase = sink while s != source: # Find the minimum value in select path _UpperCamelCase = min(__snake_case , graph[parent[s]][s] ) _UpperCamelCase = parent[s] max_flow += path_flow _UpperCamelCase = sink while v != source: _UpperCamelCase = parent[v] graph[u][v] -= path_flow graph[v][u] += path_flow _UpperCamelCase = parent[v] return max_flow _lowerCAmelCase = [ [0, 16, 13, 0, 0, 0], [0, 0, 10, 12, 0, 0], [0, 4, 0, 0, 14, 0], [0, 0, 9, 0, 0, 20], [0, 0, 0, 7, 0, 4], [0, 0, 0, 0, 0, 0], ] _lowerCAmelCase, _lowerCAmelCase = 0, 5 print(ford_fulkerson(graph, source, sink))
71
from typing import TYPE_CHECKING from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available _lowerCAmelCase = { "configuration_jukebox": [ "JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP", "JukeboxConfig", "JukeboxPriorConfig", "JukeboxVQVAEConfig", ], "tokenization_jukebox": ["JukeboxTokenizer"], } try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST", "JukeboxModel", "JukeboxPreTrainedModel", "JukeboxVQVAE", "JukeboxPrior", ] if TYPE_CHECKING: from .configuration_jukebox import ( JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP, JukeboxConfig, JukeboxPriorConfig, JukeboxVQVAEConfig, ) from .tokenization_jukebox import JukeboxTokenizer try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_jukebox import ( JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST, JukeboxModel, JukeboxPreTrainedModel, JukeboxPrior, JukeboxVQVAE, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
1
from tempfile import TemporaryDirectory from unittest import TestCase from unittest.mock import MagicMock, patch from transformers import AutoModel, TFAutoModel from transformers.onnx import FeaturesManager from transformers.testing_utils import SMALL_MODEL_IDENTIFIER, require_tf, require_torch @require_torch @require_tf class lowerCAmelCase_ ( __lowercase ): def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = SMALL_MODEL_IDENTIFIER _UpperCamelCase = '''pt''' _UpperCamelCase = '''tf''' def UpperCamelCase_ ( self : Any , _A : Tuple ): _UpperCamelCase = AutoModel.from_pretrained(self.test_model ) model_pt.save_pretrained(_A ) def UpperCamelCase_ ( self : Any , _A : Union[str, Any] ): _UpperCamelCase = TFAutoModel.from_pretrained(self.test_model , from_pt=_A ) model_tf.save_pretrained(_A ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = '''mock_framework''' # Framework provided - return whatever the user provides _UpperCamelCase = FeaturesManager.determine_framework(self.test_model , _A ) self.assertEqual(_A , _A ) # Local checkpoint and framework provided - return provided framework # PyTorch checkpoint with TemporaryDirectory() as local_pt_ckpt: self._setup_pt_ckpt(_A ) _UpperCamelCase = FeaturesManager.determine_framework(_A , _A ) self.assertEqual(_A , _A ) # TensorFlow checkpoint with TemporaryDirectory() as local_tf_ckpt: self._setup_tf_ckpt(_A ) _UpperCamelCase = FeaturesManager.determine_framework(_A , _A ) self.assertEqual(_A , _A ) def UpperCamelCase_ ( self : Optional[int] ): # PyTorch checkpoint with TemporaryDirectory() as local_pt_ckpt: self._setup_pt_ckpt(_A ) _UpperCamelCase = FeaturesManager.determine_framework(_A ) self.assertEqual(_A , self.framework_pt ) # TensorFlow checkpoint with TemporaryDirectory() as local_tf_ckpt: self._setup_tf_ckpt(_A ) _UpperCamelCase = FeaturesManager.determine_framework(_A ) self.assertEqual(_A , self.framework_tf ) # Invalid local checkpoint with TemporaryDirectory() as local_invalid_ckpt: with self.assertRaises(_A ): _UpperCamelCase = FeaturesManager.determine_framework(_A ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = MagicMock(return_value=_A ) with patch('''transformers.onnx.features.is_tf_available''' , _A ): _UpperCamelCase = FeaturesManager.determine_framework(self.test_model ) self.assertEqual(_A , self.framework_pt ) # PyTorch not in environment -> use TensorFlow _UpperCamelCase = MagicMock(return_value=_A ) with patch('''transformers.onnx.features.is_torch_available''' , _A ): _UpperCamelCase = FeaturesManager.determine_framework(self.test_model ) self.assertEqual(_A , self.framework_tf ) # Both in environment -> use PyTorch _UpperCamelCase = MagicMock(return_value=_A ) _UpperCamelCase = MagicMock(return_value=_A ) with patch('''transformers.onnx.features.is_tf_available''' , _A ), patch( '''transformers.onnx.features.is_torch_available''' , _A ): _UpperCamelCase = FeaturesManager.determine_framework(self.test_model ) self.assertEqual(_A , self.framework_pt ) # Both not in environment -> raise error _UpperCamelCase = MagicMock(return_value=_A ) _UpperCamelCase = MagicMock(return_value=_A ) with patch('''transformers.onnx.features.is_tf_available''' , _A ), patch( '''transformers.onnx.features.is_torch_available''' , _A ): with self.assertRaises(_A ): _UpperCamelCase = FeaturesManager.determine_framework(self.test_model )
71
import multiprocessing import os from typing import BinaryIO, Optional, Union import fsspec from .. import Dataset, Features, NamedSplit, config from ..formatting import query_table from ..packaged_modules.json.json import Json from ..utils import logging from ..utils.typing import NestedDataStructureLike, PathLike from .abc import AbstractDatasetReader class lowerCAmelCase_ ( __lowercase ): def __init__( self : int , _A : NestedDataStructureLike[PathLike] , _A : Optional[NamedSplit] = None , _A : Optional[Features] = None , _A : str = None , _A : bool = False , _A : bool = False , _A : Optional[str] = None , _A : Optional[int] = None , **_A : str , ): super().__init__( _A , split=_A , features=_A , cache_dir=_A , keep_in_memory=_A , streaming=_A , num_proc=_A , **_A , ) _UpperCamelCase = field _UpperCamelCase = path_or_paths if isinstance(_A , _A ) else {self.split: path_or_paths} _UpperCamelCase = Json( cache_dir=_A , data_files=_A , features=_A , field=_A , **_A , ) def UpperCamelCase_ ( self : List[str] ): # Build iterable dataset if self.streaming: _UpperCamelCase = self.builder.as_streaming_dataset(split=self.split ) # Build regular (map-style) dataset else: _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None self.builder.download_and_prepare( download_config=_A , download_mode=_A , verification_mode=_A , base_path=_A , num_proc=self.num_proc , ) _UpperCamelCase = self.builder.as_dataset( split=self.split , verification_mode=_A , in_memory=self.keep_in_memory ) return dataset class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : Dataset , _A : Union[PathLike, BinaryIO] , _A : Optional[int] = None , _A : Optional[int] = None , **_A : List[str] , ): if num_proc is not None and num_proc <= 0: raise ValueError(F"""num_proc {num_proc} must be an integer > 0.""" ) _UpperCamelCase = dataset _UpperCamelCase = path_or_buf _UpperCamelCase = batch_size if batch_size else config.DEFAULT_MAX_BATCH_SIZE _UpperCamelCase = num_proc _UpperCamelCase = '''utf-8''' _UpperCamelCase = to_json_kwargs def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = self.to_json_kwargs.pop('''path_or_buf''' , _A ) _UpperCamelCase = self.to_json_kwargs.pop('''orient''' , '''records''' ) _UpperCamelCase = self.to_json_kwargs.pop('''lines''' , True if orient == '''records''' else False ) _UpperCamelCase = self.to_json_kwargs.pop('''index''' , False if orient in ['''split''', '''table'''] else True ) _UpperCamelCase = self.to_json_kwargs.pop('''compression''' , _A ) if compression not in [None, "infer", "gzip", "bz2", "xz"]: raise NotImplementedError(F"""`datasets` currently does not support {compression} compression""" ) if isinstance(self.path_or_buf , (str, bytes, os.PathLike) ): with fsspec.open(self.path_or_buf , '''wb''' , compression=_A ) as buffer: _UpperCamelCase = self._write(file_obj=_A , orient=_A , lines=_A , index=_A , **self.to_json_kwargs ) else: if compression: raise NotImplementedError( F"""The compression parameter is not supported when writing to a buffer, but compression={compression}""" ''' was passed. Please provide a local path instead.''' ) _UpperCamelCase = self._write( file_obj=self.path_or_buf , orient=_A , lines=_A , index=_A , **self.to_json_kwargs ) return written def UpperCamelCase_ ( self : Any , _A : Optional[Any] ): _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = args _UpperCamelCase = query_table( table=self.dataset.data , key=slice(_A , offset + self.batch_size ) , indices=self.dataset._indices , ) _UpperCamelCase = batch.to_pandas().to_json( path_or_buf=_A , orient=_A , lines=_A , index=_A , **_A ) if not json_str.endswith('''\n''' ): json_str += "\n" return json_str.encode(self.encoding ) def UpperCamelCase_ ( self : int , _A : BinaryIO , _A : Dict , _A : Optional[Any] , _A : Dict , **_A : str , ): _UpperCamelCase = 0 if self.num_proc is None or self.num_proc == 1: for offset in logging.tqdm( range(0 , len(self.dataset ) , self.batch_size ) , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating json from Arrow format''' , ): _UpperCamelCase = self._batch_json((offset, orient, lines, index, to_json_kwargs) ) written += file_obj.write(_A ) else: _UpperCamelCase , _UpperCamelCase = len(self.dataset ), self.batch_size with multiprocessing.Pool(self.num_proc ) as pool: for json_str in logging.tqdm( pool.imap( self._batch_json , [(offset, orient, lines, index, to_json_kwargs) for offset in range(0 , _A , _A )] , ) , total=(num_rows // batch_size) + 1 if num_rows % batch_size else num_rows // batch_size , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating json from Arrow format''' , ): written += file_obj.write(_A ) return written
71
1
from typing import Any import numpy as np def _snake_case ( __snake_case ): return np.array_equal(__snake_case , matrix.conjugate().T ) def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = v.conjugate().T _UpperCamelCase = v_star.dot(__snake_case ) assert isinstance(__snake_case , np.ndarray ) return (v_star_dot.dot(__snake_case )) / (v_star.dot(__snake_case )) def _snake_case ( ): _UpperCamelCase = np.array([[2, 2 + 1j, 4], [2 - 1j, 3, 1j], [4, -1j, 1]] ) _UpperCamelCase = np.array([[1], [2], [3]] ) assert is_hermitian(__snake_case ), f"""{a} is not hermitian.""" print(rayleigh_quotient(__snake_case , __snake_case ) ) _UpperCamelCase = np.array([[1, 2, 4], [2, 3, -1], [4, -1, 1]] ) assert is_hermitian(__snake_case ), f"""{a} is not hermitian.""" assert rayleigh_quotient(__snake_case , __snake_case ) == float(3 ) if __name__ == "__main__": import doctest doctest.testmod() tests()
71
import enum import warnings from .. import MODEL_FOR_CAUSAL_LM_MAPPING, TF_MODEL_FOR_CAUSAL_LM_MAPPING from ..utils import add_end_docstrings, is_tf_available from .base import PIPELINE_INIT_ARGS, Pipeline if is_tf_available(): import tensorflow as tf class lowerCAmelCase_ ( enum.Enum ): UpperCAmelCase = 0 UpperCAmelCase = 1 UpperCAmelCase = 2 @add_end_docstrings(__lowercase ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "\n In 1991, the remains of Russian Tsar Nicholas II and his family (except for Alexei and Maria) are discovered. The\n voice of Nicholas's young son, Tsarevich Alexei Nikolaevich, narrates the remainder of the story. 1883 Western\n Siberia, a young Grigori Rasputin is asked by his father and a group of men to perform magic. Rasputin has a vision\n and denounces one of the men as a horse thief. Although his father initially slaps him for making such an\n accusation, Rasputin watches as the man is chased outside and beaten. Twenty years later, Rasputin sees a vision of\n the Virgin Mary, prompting him to become a priest. Rasputin quickly becomes famous, with people, even a bishop,\n begging for his blessing. <eod> </s> <eos>\n " def __init__( self : Tuple , *_A : List[str] , **_A : str ): super().__init__(*_A , **_A ) self.check_model_type( TF_MODEL_FOR_CAUSAL_LM_MAPPING if self.framework == '''tf''' else MODEL_FOR_CAUSAL_LM_MAPPING ) if "prefix" not in self._preprocess_params: # This is very specific. The logic is quite complex and needs to be done # as a "default". # It also defines both some preprocess_kwargs and generate_kwargs # which is why we cannot put them in their respective methods. _UpperCamelCase = None if self.model.config.prefix is not None: _UpperCamelCase = self.model.config.prefix if prefix is None and self.model.__class__.__name__ in [ "XLNetLMHeadModel", "TransfoXLLMHeadModel", "TFXLNetLMHeadModel", "TFTransfoXLLMHeadModel", ]: # For XLNet and TransformerXL we add an article to the prompt to give more state to the model. _UpperCamelCase = self.XL_PREFIX if prefix is not None: # Recalculate some generate_kwargs linked to prefix. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = self._sanitize_parameters(prefix=_A , **self._forward_params ) _UpperCamelCase = {**self._preprocess_params, **preprocess_params} _UpperCamelCase = {**self._forward_params, **forward_params} def UpperCamelCase_ ( self : Dict , _A : Optional[int]=None , _A : Any=None , _A : Optional[int]=None , _A : List[str]=None , _A : List[Any]=None , _A : int=None , _A : Tuple=None , _A : Optional[Any]=None , **_A : Optional[int] , ): _UpperCamelCase = {} if prefix is not None: _UpperCamelCase = prefix if prefix: _UpperCamelCase = self.tokenizer( _A , padding=_A , add_special_tokens=_A , return_tensors=self.framework ) _UpperCamelCase = prefix_inputs['''input_ids'''].shape[-1] if handle_long_generation is not None: if handle_long_generation not in {"hole"}: raise ValueError( F"""{handle_long_generation} is not a valid value for `handle_long_generation` parameter expected""" ''' [None, \'hole\']''' ) _UpperCamelCase = handle_long_generation preprocess_params.update(_A ) _UpperCamelCase = generate_kwargs _UpperCamelCase = {} if return_full_text is not None and return_type is None: if return_text is not None: raise ValueError('''`return_text` is mutually exclusive with `return_full_text`''' ) if return_tensors is not None: raise ValueError('''`return_full_text` is mutually exclusive with `return_tensors`''' ) _UpperCamelCase = ReturnType.FULL_TEXT if return_full_text else ReturnType.NEW_TEXT if return_tensors is not None and return_type is None: if return_text is not None: raise ValueError('''`return_text` is mutually exclusive with `return_tensors`''' ) _UpperCamelCase = ReturnType.TENSORS if return_type is not None: _UpperCamelCase = return_type if clean_up_tokenization_spaces is not None: _UpperCamelCase = clean_up_tokenization_spaces if stop_sequence is not None: _UpperCamelCase = self.tokenizer.encode(_A , add_special_tokens=_A ) if len(_A ) > 1: warnings.warn( '''Stopping on a multiple token sequence is not yet supported on transformers. The first token of''' ''' the stop sequence will be used as the stop sequence string in the interim.''' ) _UpperCamelCase = stop_sequence_ids[0] return preprocess_params, forward_params, postprocess_params def UpperCamelCase_ ( self : int , *_A : Union[str, Any] , **_A : Union[str, Any] ): # Parse arguments if self.model.__class__.__name__ in ["TransfoXLLMHeadModel"]: kwargs.update({'''add_space_before_punct_symbol''': True} ) return super()._parse_and_tokenize(*_A , **_A ) def __call__( self : List[str] , _A : str , **_A : Any ): return super().__call__(_A , **_A ) def UpperCamelCase_ ( self : Optional[Any] , _A : List[str] , _A : int="" , _A : Optional[Any]=None , **_A : Optional[Any] ): _UpperCamelCase = self.tokenizer( prefix + prompt_text , padding=_A , add_special_tokens=_A , return_tensors=self.framework ) _UpperCamelCase = prompt_text if handle_long_generation == "hole": _UpperCamelCase = inputs['''input_ids'''].shape[-1] if "max_new_tokens" in generate_kwargs: _UpperCamelCase = generate_kwargs['''max_new_tokens'''] else: _UpperCamelCase = generate_kwargs.get('''max_length''' , self.model.config.max_length ) - cur_len if new_tokens < 0: raise ValueError('''We cannot infer how many new tokens are expected''' ) if cur_len + new_tokens > self.tokenizer.model_max_length: _UpperCamelCase = self.tokenizer.model_max_length - new_tokens if keep_length <= 0: raise ValueError( '''We cannot use `hole` to handle this generation the number of desired tokens exceeds the''' ''' models max length''' ) _UpperCamelCase = inputs['''input_ids'''][:, -keep_length:] if "attention_mask" in inputs: _UpperCamelCase = inputs['''attention_mask'''][:, -keep_length:] return inputs def UpperCamelCase_ ( self : Dict , _A : Optional[int] , **_A : str ): _UpperCamelCase = model_inputs['''input_ids'''] _UpperCamelCase = model_inputs.get('''attention_mask''' , _A ) # Allow empty prompts if input_ids.shape[1] == 0: _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = 1 else: _UpperCamelCase = input_ids.shape[0] _UpperCamelCase = model_inputs.pop('''prompt_text''' ) # If there is a prefix, we may need to adjust the generation length. Do so without permanently modifying # generate_kwargs, as some of the parameterization may come from the initialization of the pipeline. _UpperCamelCase = generate_kwargs.pop('''prefix_length''' , 0 ) if prefix_length > 0: _UpperCamelCase = '''max_new_tokens''' in generate_kwargs or ( '''generation_config''' in generate_kwargs and generate_kwargs['''generation_config'''].max_new_tokens is not None ) if not has_max_new_tokens: _UpperCamelCase = generate_kwargs.get('''max_length''' ) or self.model.config.max_length generate_kwargs["max_length"] += prefix_length _UpperCamelCase = '''min_new_tokens''' in generate_kwargs or ( '''generation_config''' in generate_kwargs and generate_kwargs['''generation_config'''].min_new_tokens is not None ) if not has_min_new_tokens and "min_length" in generate_kwargs: generate_kwargs["min_length"] += prefix_length # BS x SL _UpperCamelCase = self.model.generate(input_ids=_A , attention_mask=_A , **_A ) _UpperCamelCase = generated_sequence.shape[0] if self.framework == "pt": _UpperCamelCase = generated_sequence.reshape(_A , out_b // in_b , *generated_sequence.shape[1:] ) elif self.framework == "tf": _UpperCamelCase = tf.reshape(_A , (in_b, out_b // in_b, *generated_sequence.shape[1:]) ) return {"generated_sequence": generated_sequence, "input_ids": input_ids, "prompt_text": prompt_text} def UpperCamelCase_ ( self : List[str] , _A : Dict , _A : Optional[Any]=ReturnType.FULL_TEXT , _A : Dict=True ): _UpperCamelCase = model_outputs['''generated_sequence'''][0] _UpperCamelCase = model_outputs['''input_ids'''] _UpperCamelCase = model_outputs['''prompt_text'''] _UpperCamelCase = generated_sequence.numpy().tolist() _UpperCamelCase = [] for sequence in generated_sequence: if return_type == ReturnType.TENSORS: _UpperCamelCase = {'''generated_token_ids''': sequence} elif return_type in {ReturnType.NEW_TEXT, ReturnType.FULL_TEXT}: # Decode text _UpperCamelCase = self.tokenizer.decode( _A , skip_special_tokens=_A , clean_up_tokenization_spaces=_A , ) # Remove PADDING prompt of the sequence if XLNet or Transfo-XL model is used if input_ids is None: _UpperCamelCase = 0 else: _UpperCamelCase = len( self.tokenizer.decode( input_ids[0] , skip_special_tokens=_A , clean_up_tokenization_spaces=_A , ) ) if return_type == ReturnType.FULL_TEXT: _UpperCamelCase = prompt_text + text[prompt_length:] else: _UpperCamelCase = text[prompt_length:] _UpperCamelCase = {'''generated_text''': all_text} records.append(_A ) return records
71
1
from ..utils import DummyObject, requires_backends class lowerCAmelCase_ ( metaclass=__lowercase ): UpperCAmelCase = ["keras_nlp"] def __init__( self : Any , *_A : Dict , **_A : List[str] ): requires_backends(self , ['''keras_nlp'''] )
71
import unittest from transformers import is_torch_available from transformers.testing_utils import require_sentencepiece, require_tokenizers, require_torch, slow, torch_device if is_torch_available(): from transformers import AutoModelForSeqaSeqLM, AutoTokenizer @require_torch @require_sentencepiece @require_tokenizers class lowerCAmelCase_ ( unittest.TestCase ): @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = AutoModelForSeqaSeqLM.from_pretrained('''google/mt5-small''' , return_dict=_A ).to(_A ) _UpperCamelCase = AutoTokenizer.from_pretrained('''google/mt5-small''' ) _UpperCamelCase = tokenizer('''Hello there''' , return_tensors='''pt''' ).input_ids _UpperCamelCase = tokenizer('''Hi I am''' , return_tensors='''pt''' ).input_ids _UpperCamelCase = model(input_ids.to(_A ) , labels=labels.to(_A ) ).loss _UpperCamelCase = -(labels.shape[-1] * loss.item()) _UpperCamelCase = -84.9127 self.assertTrue(abs(mtf_score - EXPECTED_SCORE ) < 1e-4 )
71
1
from __future__ import annotations class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : str , _A : str ): _UpperCamelCase , _UpperCamelCase = text, pattern _UpperCamelCase , _UpperCamelCase = len(_A ), len(_A ) def UpperCamelCase_ ( self : List[Any] , _A : str ): for i in range(self.patLen - 1 , -1 , -1 ): if char == self.pattern[i]: return i return -1 def UpperCamelCase_ ( self : Union[str, Any] , _A : int ): for i in range(self.patLen - 1 , -1 , -1 ): if self.pattern[i] != self.text[current_pos + i]: return current_pos + i return -1 def UpperCamelCase_ ( self : List[str] ): # searches pattern in text and returns index positions _UpperCamelCase = [] for i in range(self.textLen - self.patLen + 1 ): _UpperCamelCase = self.mismatch_in_text(_A ) if mismatch_index == -1: positions.append(_A ) else: _UpperCamelCase = self.match_in_pattern(self.text[mismatch_index] ) _UpperCamelCase = ( mismatch_index - match_index ) # shifting index lgtm [py/multiple-definition] return positions _lowerCAmelCase = "ABAABA" _lowerCAmelCase = "AB" _lowerCAmelCase = BoyerMooreSearch(text, pattern) _lowerCAmelCase = bms.bad_character_heuristic() if len(positions) == 0: print("No match found") else: print("Pattern found in following positions: ") print(positions)
71
import logging import os import sys from dataclasses import dataclass, field from typing import Optional from seqaseq_trainer import SeqaSeqTrainer from seqaseq_training_args import SeqaSeqTrainingArguments import transformers from transformers import ( AutoConfig, AutoModelForSeqaSeqLM, AutoTokenizer, HfArgumentParser, MBartTokenizer, MBartTokenizerFast, set_seed, ) from transformers.trainer_utils import EvaluationStrategy, is_main_process from transformers.training_args import ParallelMode from utils import ( SeqaSeqDataCollator, SeqaSeqDataset, assert_all_frozen, build_compute_metrics_fn, check_output_dir, freeze_embeds, freeze_params, lmap, save_json, use_task_specific_params, write_txt_file, ) _lowerCAmelCase = logging.getLogger(__name__) @dataclass class lowerCAmelCase_ : UpperCAmelCase = field( metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Pretrained config name or path if not the same as model_name"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"}, ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Whether tp freeze the encoder."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Whether to freeze the embeddings."} ) @dataclass class lowerCAmelCase_ : UpperCAmelCase = field( metadata={"help": "The input data dir. Should contain the .tsv files (or other data files) for the task."} ) UpperCAmelCase = field( default="summarization", metadata={"help": "Task name, summarization (or summarization_{dataset} for pegasus) or translation"}, ) UpperCAmelCase = field( default=1024, metadata={ "help": ( "The maximum total input sequence length after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field( default=128, metadata={ "help": ( "The maximum total sequence length for target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field( default=142, metadata={ "help": ( "The maximum total sequence length for validation target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded. " "This argument is also used to override the ``max_length`` param of ``model.generate``, which is used " "during ``evaluate`` and ``predict``." ) }, ) UpperCAmelCase = field( default=142, metadata={ "help": ( "The maximum total sequence length for test target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field(default=-1, metadata={"help": "# training examples. -1 means use all."} ) UpperCAmelCase = field(default=-1, metadata={"help": "# validation examples. -1 means use all."} ) UpperCAmelCase = field(default=-1, metadata={"help": "# test examples. -1 means use all."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Source language id for translation."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Target language id for translation."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "# num_beams to use for evaluation."} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "If only pad tokens should be ignored. This assumes that `config.pad_token_id` is defined."}, ) def _snake_case ( __snake_case , __snake_case , __snake_case ): logger.info(f"""***** {split} metrics *****""" ) for key in sorted(metrics.keys() ): logger.info(f""" {key} = {metrics[key]}""" ) save_json(__snake_case , os.path.join(__snake_case , f"""{split}_results.json""" ) ) def _snake_case ( ): # See all possible arguments in src/transformers/training_args.py # or by passing the --help flag to this script. # We now keep distinct sets of args, for a cleaner separation of concerns. _UpperCamelCase = HfArgumentParser((ModelArguments, DataTrainingArguments, SeqaSeqTrainingArguments) ) if len(sys.argv ) == 2 and sys.argv[1].endswith('''.json''' ): # If we pass only one argument to the script and it's the path to a json file, # let's parse it to get our arguments. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1] ) ) else: _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = parser.parse_args_into_dataclasses() check_output_dir(__snake_case ) # Setup logging logging.basicConfig( format='''%(asctime)s - %(levelname)s - %(name)s - %(message)s''' , datefmt='''%m/%d/%Y %H:%M:%S''' , level=logging.INFO if training_args.local_rank in [-1, 0] else logging.WARN , ) logger.warning( '''Process rank: %s, device: %s, n_gpu: %s, distributed training: %s, 16-bits training: %s''' , training_args.local_rank , training_args.device , training_args.n_gpu , bool(training_args.parallel_mode == ParallelMode.DISTRIBUTED ) , training_args.fpaa , ) transformers.utils.logging.enable_default_handler() transformers.utils.logging.enable_explicit_format() # Set the verbosity to info of the Transformers logger (on main process only): if is_main_process(training_args.local_rank ): transformers.utils.logging.set_verbosity_info() logger.info('''Training/evaluation parameters %s''' , __snake_case ) # Set seed set_seed(training_args.seed ) # Load pretrained model and tokenizer # # Distributed training: # The .from_pretrained methods guarantee that only one local process can concurrently # download model & vocab. _UpperCamelCase = AutoConfig.from_pretrained( model_args.config_name if model_args.config_name else model_args.model_name_or_path , cache_dir=model_args.cache_dir , ) _UpperCamelCase = ('''encoder_layerdrop''', '''decoder_layerdrop''', '''dropout''', '''attention_dropout''') for p in extra_model_params: if getattr(__snake_case , __snake_case , __snake_case ): assert hasattr(__snake_case , __snake_case ), f"""({config.__class__.__name__}) doesn't have a `{p}` attribute""" setattr(__snake_case , __snake_case , getattr(__snake_case , __snake_case ) ) _UpperCamelCase = AutoTokenizer.from_pretrained( model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path , cache_dir=model_args.cache_dir , ) _UpperCamelCase = AutoModelForSeqaSeqLM.from_pretrained( model_args.model_name_or_path , from_tf='''.ckpt''' in model_args.model_name_or_path , config=__snake_case , cache_dir=model_args.cache_dir , ) # use task specific params use_task_specific_params(__snake_case , data_args.task ) # set num_beams for evaluation if data_args.eval_beams is None: _UpperCamelCase = model.config.num_beams # set decoder_start_token_id for MBart if model.config.decoder_start_token_id is None and isinstance(__snake_case , (MBartTokenizer, MBartTokenizerFast) ): assert ( data_args.tgt_lang is not None and data_args.src_lang is not None ), "mBart requires --tgt_lang and --src_lang" if isinstance(__snake_case , __snake_case ): _UpperCamelCase = tokenizer.lang_code_to_id[data_args.tgt_lang] else: _UpperCamelCase = tokenizer.convert_tokens_to_ids(data_args.tgt_lang ) if model_args.freeze_embeds: freeze_embeds(__snake_case ) if model_args.freeze_encoder: freeze_params(model.get_encoder() ) assert_all_frozen(model.get_encoder() ) _UpperCamelCase = SeqaSeqDataset # Get datasets _UpperCamelCase = ( dataset_class( __snake_case , type_path='''train''' , data_dir=data_args.data_dir , n_obs=data_args.n_train , max_target_length=data_args.max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_train else None ) _UpperCamelCase = ( dataset_class( __snake_case , type_path='''val''' , data_dir=data_args.data_dir , n_obs=data_args.n_val , max_target_length=data_args.val_max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_eval or training_args.evaluation_strategy != EvaluationStrategy.NO else None ) _UpperCamelCase = ( dataset_class( __snake_case , type_path='''test''' , data_dir=data_args.data_dir , n_obs=data_args.n_test , max_target_length=data_args.test_max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_predict else None ) # Initialize our Trainer _UpperCamelCase = ( build_compute_metrics_fn(data_args.task , __snake_case ) if training_args.predict_with_generate else None ) _UpperCamelCase = SeqaSeqTrainer( model=__snake_case , args=__snake_case , data_args=__snake_case , train_dataset=__snake_case , eval_dataset=__snake_case , data_collator=SeqaSeqDataCollator( __snake_case , __snake_case , model.config.decoder_start_token_id , training_args.tpu_num_cores ) , compute_metrics=__snake_case , tokenizer=__snake_case , ) _UpperCamelCase = {} # Training if training_args.do_train: logger.info('''*** Train ***''' ) _UpperCamelCase = trainer.train( model_path=model_args.model_name_or_path if os.path.isdir(model_args.model_name_or_path ) else None ) _UpperCamelCase = train_result.metrics _UpperCamelCase = data_args.n_train trainer.save_model() # this also saves the tokenizer if trainer.is_world_process_zero(): handle_metrics('''train''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) # Need to save the state, since Trainer.save_model saves only the tokenizer with the model trainer.state.save_to_json(os.path.join(training_args.output_dir , '''trainer_state.json''' ) ) # For convenience, we also re-save the tokenizer to the same directory, # so that you can share your model easily on huggingface.co/models =) tokenizer.save_pretrained(training_args.output_dir ) # Evaluation if training_args.do_eval: logger.info('''*** Evaluate ***''' ) _UpperCamelCase = trainer.evaluate(metric_key_prefix='''val''' ) _UpperCamelCase = data_args.n_val _UpperCamelCase = round(metrics['''val_loss'''] , 4 ) if trainer.is_world_process_zero(): handle_metrics('''val''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) if training_args.do_predict: logger.info('''*** Predict ***''' ) _UpperCamelCase = trainer.predict(test_dataset=__snake_case , metric_key_prefix='''test''' ) _UpperCamelCase = test_output.metrics _UpperCamelCase = data_args.n_test if trainer.is_world_process_zero(): _UpperCamelCase = round(metrics['''test_loss'''] , 4 ) handle_metrics('''test''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) if training_args.predict_with_generate: _UpperCamelCase = tokenizer.batch_decode( test_output.predictions , skip_special_tokens=__snake_case , clean_up_tokenization_spaces=__snake_case ) _UpperCamelCase = lmap(str.strip , __snake_case ) write_txt_file(__snake_case , os.path.join(training_args.output_dir , '''test_generations.txt''' ) ) if trainer.is_world_process_zero(): save_json(__snake_case , os.path.join(training_args.output_dir , '''all_results.json''' ) ) return all_metrics def _snake_case ( __snake_case ): # For xla_spawn (TPUs) main() if __name__ == "__main__": main()
71
1
import argparse import json from collections import OrderedDict from pathlib import Path import requests import torch from huggingface_hub import hf_hub_download from PIL import Image from transformers import ( SegformerConfig, SegformerForImageClassification, SegformerForSemanticSegmentation, SegformerImageProcessor, ) from transformers.utils import logging logging.set_verbosity_info() _lowerCAmelCase = logging.get_logger(__name__) def _snake_case ( __snake_case , __snake_case=False ): _UpperCamelCase = OrderedDict() for key, value in state_dict.items(): if encoder_only and not key.startswith('''head''' ): _UpperCamelCase = '''segformer.encoder.''' + key if key.startswith('''backbone''' ): _UpperCamelCase = key.replace('''backbone''' , '''segformer.encoder''' ) if "patch_embed" in key: # replace for example patch_embed1 by patch_embeddings.0 _UpperCamelCase = key[key.find('''patch_embed''' ) + len('''patch_embed''' )] _UpperCamelCase = key.replace(f"""patch_embed{idx}""" , f"""patch_embeddings.{int(__snake_case )-1}""" ) if "norm" in key: _UpperCamelCase = key.replace('''norm''' , '''layer_norm''' ) if "segformer.encoder.layer_norm" in key: # replace for example layer_norm1 by layer_norm.0 _UpperCamelCase = key[key.find('''segformer.encoder.layer_norm''' ) + len('''segformer.encoder.layer_norm''' )] _UpperCamelCase = key.replace(f"""layer_norm{idx}""" , f"""layer_norm.{int(__snake_case )-1}""" ) if "layer_norm1" in key: _UpperCamelCase = key.replace('''layer_norm1''' , '''layer_norm_1''' ) if "layer_norm2" in key: _UpperCamelCase = key.replace('''layer_norm2''' , '''layer_norm_2''' ) if "block" in key: # replace for example block1 by block.0 _UpperCamelCase = key[key.find('''block''' ) + len('''block''' )] _UpperCamelCase = key.replace(f"""block{idx}""" , f"""block.{int(__snake_case )-1}""" ) if "attn.q" in key: _UpperCamelCase = key.replace('''attn.q''' , '''attention.self.query''' ) if "attn.proj" in key: _UpperCamelCase = key.replace('''attn.proj''' , '''attention.output.dense''' ) if "attn" in key: _UpperCamelCase = key.replace('''attn''' , '''attention.self''' ) if "fc1" in key: _UpperCamelCase = key.replace('''fc1''' , '''dense1''' ) if "fc2" in key: _UpperCamelCase = key.replace('''fc2''' , '''dense2''' ) if "linear_pred" in key: _UpperCamelCase = key.replace('''linear_pred''' , '''classifier''' ) if "linear_fuse" in key: _UpperCamelCase = key.replace('''linear_fuse.conv''' , '''linear_fuse''' ) _UpperCamelCase = key.replace('''linear_fuse.bn''' , '''batch_norm''' ) if "linear_c" in key: # replace for example linear_c4 by linear_c.3 _UpperCamelCase = key[key.find('''linear_c''' ) + len('''linear_c''' )] _UpperCamelCase = key.replace(f"""linear_c{idx}""" , f"""linear_c.{int(__snake_case )-1}""" ) if key.startswith('''head''' ): _UpperCamelCase = key.replace('''head''' , '''classifier''' ) _UpperCamelCase = value return new_state_dict def _snake_case ( __snake_case , __snake_case ): # for each of the encoder blocks: for i in range(config.num_encoder_blocks ): for j in range(config.depths[i] ): # read in weights + bias of keys and values (which is a single matrix in the original implementation) _UpperCamelCase = state_dict.pop(f"""segformer.encoder.block.{i}.{j}.attention.self.kv.weight""" ) _UpperCamelCase = state_dict.pop(f"""segformer.encoder.block.{i}.{j}.attention.self.kv.bias""" ) # next, add keys and values (in that order) to the state dict _UpperCamelCase = kv_weight[ : config.hidden_sizes[i], : ] _UpperCamelCase = kv_bias[: config.hidden_sizes[i]] _UpperCamelCase = kv_weight[ config.hidden_sizes[i] :, : ] _UpperCamelCase = kv_bias[ config.hidden_sizes[i] : ] def _snake_case ( ): _UpperCamelCase = '''http://images.cocodataset.org/val2017/000000039769.jpg''' _UpperCamelCase = Image.open(requests.get(__snake_case , stream=__snake_case ).raw ) return image @torch.no_grad() def _snake_case ( __snake_case , __snake_case , __snake_case ): _UpperCamelCase = SegformerConfig() _UpperCamelCase = False # set attributes based on model_name _UpperCamelCase = '''huggingface/label-files''' if "segformer" in model_name: _UpperCamelCase = model_name[len('''segformer.''' ) : len('''segformer.''' ) + 2] if "ade" in model_name: _UpperCamelCase = 150 _UpperCamelCase = '''ade20k-id2label.json''' _UpperCamelCase = (1, 150, 128, 128) elif "city" in model_name: _UpperCamelCase = 19 _UpperCamelCase = '''cityscapes-id2label.json''' _UpperCamelCase = (1, 19, 128, 128) else: raise ValueError(f"""Model {model_name} not supported""" ) elif "mit" in model_name: _UpperCamelCase = True _UpperCamelCase = model_name[4:6] _UpperCamelCase = 1000 _UpperCamelCase = '''imagenet-1k-id2label.json''' _UpperCamelCase = (1, 1000) else: raise ValueError(f"""Model {model_name} not supported""" ) # set config attributes _UpperCamelCase = json.load(open(hf_hub_download(__snake_case , __snake_case , repo_type='''dataset''' ) , '''r''' ) ) _UpperCamelCase = {int(__snake_case ): v for k, v in idalabel.items()} _UpperCamelCase = idalabel _UpperCamelCase = {v: k for k, v in idalabel.items()} if size == "b0": pass elif size == "b1": _UpperCamelCase = [64, 128, 320, 512] _UpperCamelCase = 256 elif size == "b2": _UpperCamelCase = [64, 128, 320, 512] _UpperCamelCase = 768 _UpperCamelCase = [3, 4, 6, 3] elif size == "b3": _UpperCamelCase = [64, 128, 320, 512] _UpperCamelCase = 768 _UpperCamelCase = [3, 4, 18, 3] elif size == "b4": _UpperCamelCase = [64, 128, 320, 512] _UpperCamelCase = 768 _UpperCamelCase = [3, 8, 27, 3] elif size == "b5": _UpperCamelCase = [64, 128, 320, 512] _UpperCamelCase = 768 _UpperCamelCase = [3, 6, 40, 3] else: raise ValueError(f"""Size {size} not supported""" ) # load image processor (only resize + normalize) _UpperCamelCase = SegformerImageProcessor( image_scale=(512, 512) , keep_ratio=__snake_case , align=__snake_case , do_random_crop=__snake_case ) # prepare image _UpperCamelCase = prepare_img() _UpperCamelCase = image_processor(images=__snake_case , return_tensors='''pt''' ).pixel_values logger.info(f"""Converting model {model_name}...""" ) # load original state dict if encoder_only: _UpperCamelCase = torch.load(__snake_case , map_location=torch.device('''cpu''' ) ) else: _UpperCamelCase = torch.load(__snake_case , map_location=torch.device('''cpu''' ) )['''state_dict'''] # rename keys _UpperCamelCase = rename_keys(__snake_case , encoder_only=__snake_case ) if not encoder_only: del state_dict["decode_head.conv_seg.weight"] del state_dict["decode_head.conv_seg.bias"] # key and value matrices need special treatment read_in_k_v(__snake_case , __snake_case ) # create HuggingFace model and load state dict if encoder_only: _UpperCamelCase = False _UpperCamelCase = SegformerForImageClassification(__snake_case ) else: _UpperCamelCase = SegformerForSemanticSegmentation(__snake_case ) model.load_state_dict(__snake_case ) model.eval() # forward pass _UpperCamelCase = model(__snake_case ) _UpperCamelCase = outputs.logits # set expected_slice based on model name # ADE20k checkpoints if model_name == "segformer.b0.512x512.ade.160k": _UpperCamelCase = torch.tensor( [ [[-4.6310, -5.5232, -6.2356], [-5.1921, -6.1444, -6.5996], [-5.4424, -6.2790, -6.7574]], [[-12.1391, -13.3122, -13.9554], [-12.8732, -13.9352, -14.3563], [-12.9438, -13.8226, -14.2513]], [[-12.5134, -13.4686, -14.4915], [-12.8669, -14.4343, -14.7758], [-13.2523, -14.5819, -15.0694]], ] ) elif model_name == "segformer.b1.512x512.ade.160k": _UpperCamelCase = torch.tensor( [ [[-7.5820, -8.7231, -8.3215], [-8.0600, -10.3529, -10.0304], [-7.5208, -9.4103, -9.6239]], [[-12.6918, -13.8994, -13.7137], [-13.3196, -15.7523, -15.4789], [-12.9343, -14.8757, -14.9689]], [[-11.1911, -11.9421, -11.3243], [-11.3342, -13.6839, -13.3581], [-10.3909, -12.1832, -12.4858]], ] ) elif model_name == "segformer.b2.512x512.ade.160k": _UpperCamelCase = torch.tensor( [ [[-11.8173, -14.3850, -16.3128], [-14.5648, -16.5804, -18.6568], [-14.7223, -15.7387, -18.4218]], [[-15.7290, -17.9171, -19.4423], [-18.3105, -19.9448, -21.4661], [-17.9296, -18.6497, -20.7910]], [[-15.0783, -17.0336, -18.2789], [-16.8771, -18.6870, -20.1612], [-16.2454, -17.1426, -19.5055]], ] ) elif model_name == "segformer.b3.512x512.ade.160k": _UpperCamelCase = torch.tensor( [ [[-9.0878, -10.2081, -10.1891], [-9.3144, -10.7941, -10.9843], [-9.2294, -10.3855, -10.5704]], [[-12.2316, -13.9068, -13.6102], [-12.9161, -14.3702, -14.3235], [-12.5233, -13.7174, -13.7932]], [[-14.6275, -15.2490, -14.9727], [-14.3400, -15.9687, -16.2827], [-14.1484, -15.4033, -15.8937]], ] ) elif model_name == "segformer.b4.512x512.ade.160k": _UpperCamelCase = torch.tensor( [ [[-12.3144, -13.2447, -14.0802], [-13.3614, -14.5816, -15.6117], [-13.3340, -14.4433, -16.2219]], [[-19.2781, -20.4128, -20.7506], [-20.6153, -21.6566, -22.0998], [-19.9800, -21.0430, -22.1494]], [[-18.8739, -19.7804, -21.1834], [-20.1233, -21.6765, -23.2944], [-20.0315, -21.2641, -23.6944]], ] ) elif model_name == "segformer.b5.640x640.ade.160k": _UpperCamelCase = torch.tensor( [ [[-9.5524, -12.0835, -11.7348], [-10.5229, -13.6446, -14.5662], [-9.5842, -12.8851, -13.9414]], [[-15.3432, -17.5323, -17.0818], [-16.3330, -18.9255, -19.2101], [-15.1340, -17.7848, -18.3971]], [[-12.6072, -14.9486, -14.6631], [-13.7629, -17.0907, -17.7745], [-12.7899, -16.1695, -17.1671]], ] ) # Cityscapes checkpoints elif model_name == "segformer.b0.1024x1024.city.160k": _UpperCamelCase = torch.tensor( [ [[-11.9295, -13.4057, -14.8106], [-13.3431, -14.8179, -15.3781], [-14.2836, -15.5942, -16.1588]], [[-11.4906, -12.8067, -13.6564], [-13.1189, -14.0500, -14.1543], [-13.8748, -14.5136, -14.8789]], [[0.5374, 0.1067, -0.4742], [0.1141, -0.2255, -0.7099], [-0.3000, -0.5924, -1.3105]], ] ) elif model_name == "segformer.b0.512x1024.city.160k": _UpperCamelCase = torch.tensor( [ [[-7.8217, -9.8767, -10.1717], [-9.4438, -10.9058, -11.4047], [-9.7939, -12.3495, -12.1079]], [[-7.1514, -9.5336, -10.0860], [-9.7776, -11.6822, -11.8439], [-10.1411, -12.7655, -12.8972]], [[0.3021, 0.0805, -0.2310], [-0.0328, -0.1605, -0.2714], [-0.1408, -0.5477, -0.6976]], ] ) elif model_name == "segformer.b0.640x1280.city.160k": _UpperCamelCase = torch.tensor( [ [ [-1.1372E01, -1.2787E01, -1.3477E01], [-1.2536E01, -1.4194E01, -1.4409E01], [-1.3217E01, -1.4888E01, -1.5327E01], ], [ [-1.4791E01, -1.7122E01, -1.8277E01], [-1.7163E01, -1.9192E01, -1.9533E01], [-1.7897E01, -1.9991E01, -2.0315E01], ], [ [7.6723E-01, 4.1921E-01, -7.7878E-02], [4.7772E-01, 9.5557E-03, -2.8082E-01], [3.6032E-01, -2.4826E-01, -5.1168E-01], ], ] ) elif model_name == "segformer.b0.768x768.city.160k": _UpperCamelCase = torch.tensor( [ [[-9.4959, -11.3087, -11.7479], [-11.0025, -12.6540, -12.3319], [-11.4064, -13.0487, -12.9905]], [[-9.8905, -11.3084, -12.0854], [-11.1726, -12.7698, -12.9583], [-11.5985, -13.3278, -14.1774]], [[0.2213, 0.0192, -0.2466], [-0.1731, -0.4213, -0.4874], [-0.3126, -0.6541, -1.1389]], ] ) elif model_name == "segformer.b1.1024x1024.city.160k": _UpperCamelCase = torch.tensor( [ [[-13.5748, -13.9111, -12.6500], [-14.3500, -15.3683, -14.2328], [-14.7532, -16.0424, -15.6087]], [[-17.1651, -15.8725, -12.9653], [-17.2580, -17.3718, -14.8223], [-16.6058, -16.8783, -16.7452]], [[-3.6456, -3.0209, -1.4203], [-3.0797, -3.1959, -2.0000], [-1.8757, -1.9217, -1.6997]], ] ) elif model_name == "segformer.b2.1024x1024.city.160k": _UpperCamelCase = torch.tensor( [ [[-16.0976, -16.4856, -17.3962], [-16.6234, -19.0342, -19.7685], [-16.0900, -18.0661, -19.1180]], [[-18.4750, -18.8488, -19.5074], [-19.4030, -22.1570, -22.5977], [-19.1191, -20.8486, -22.3783]], [[-4.5178, -5.5037, -6.5109], [-5.0884, -7.2174, -8.0334], [-4.4156, -5.8117, -7.2970]], ] ) elif model_name == "segformer.b3.1024x1024.city.160k": _UpperCamelCase = torch.tensor( [ [[-14.2081, -14.4732, -14.1977], [-14.5867, -16.4423, -16.6356], [-13.4441, -14.9685, -16.8696]], [[-14.4576, -14.7073, -15.0451], [-15.0816, -17.6237, -17.9873], [-14.4213, -16.0199, -18.5992]], [[-4.7349, -4.9588, -5.0966], [-4.3210, -6.9325, -7.2591], [-3.4312, -4.7484, -7.1917]], ] ) elif model_name == "segformer.b4.1024x1024.city.160k": _UpperCamelCase = torch.tensor( [ [[-11.7737, -11.9526, -11.3273], [-13.6692, -14.4574, -13.8878], [-13.8937, -14.6924, -15.9345]], [[-14.6706, -14.5330, -14.1306], [-16.1502, -16.8180, -16.4269], [-16.8338, -17.8939, -20.1746]], [[1.0491, 0.8289, 1.0310], [1.1044, 0.5219, 0.8055], [1.0899, 0.6926, 0.5590]], ] ) elif model_name == "segformer.b5.1024x1024.city.160k": _UpperCamelCase = torch.tensor( [ [[-12.5641, -13.4777, -13.0684], [-13.9587, -15.8983, -16.6557], [-13.3109, -15.7350, -16.3141]], [[-14.7074, -15.4352, -14.5944], [-16.6353, -18.1663, -18.6120], [-15.1702, -18.0329, -18.1547]], [[-1.7990, -2.0951, -1.7784], [-2.6397, -3.8245, -3.9686], [-1.5264, -2.8126, -2.9316]], ] ) else: _UpperCamelCase = logits.argmax(-1 ).item() print('''Predicted class:''' , model.config.idalabel[predicted_class_idx] ) # verify logits if not encoder_only: assert logits.shape == expected_shape assert torch.allclose(logits[0, :3, :3, :3] , __snake_case , atol=1E-2 ) # finally, save model and image processor logger.info(f"""Saving PyTorch model and image processor to {pytorch_dump_folder_path}...""" ) Path(__snake_case ).mkdir(exist_ok=__snake_case ) model.save_pretrained(__snake_case ) image_processor.save_pretrained(__snake_case ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() parser.add_argument( "--model_name", default="segformer.b0.512x512.ade.160k", type=str, help="Name of the model you'd like to convert.", ) parser.add_argument( "--checkpoint_path", default=None, type=str, help="Path to the original PyTorch checkpoint (.pth file)." ) parser.add_argument( "--pytorch_dump_folder_path", default=None, type=str, help="Path to the folder to output PyTorch model." ) _lowerCAmelCase = parser.parse_args() convert_segformer_checkpoint(args.model_name, args.checkpoint_path, args.pytorch_dump_folder_path)
71
from __future__ import annotations import typing from collections import Counter def _snake_case ( __snake_case ): _UpperCamelCase = Counter() for base in range(1 , max_perimeter + 1 ): for perpendicular in range(__snake_case , max_perimeter + 1 ): _UpperCamelCase = (base * base + perpendicular * perpendicular) ** 0.5 if hypotenuse == int(__snake_case ): _UpperCamelCase = int(base + perpendicular + hypotenuse ) if perimeter > max_perimeter: continue triplets[perimeter] += 1 return triplets def _snake_case ( __snake_case = 1000 ): _UpperCamelCase = pythagorean_triple(__snake_case ) return triplets.most_common(1 )[0][0] if __name__ == "__main__": print(f'Perimeter {solution()} has maximum solutions')
71
1
from __future__ import annotations import math class lowerCAmelCase_ : def __init__( self : int , _A : int ): _UpperCamelCase = size # approximate the overall size of segment tree with given value _UpperCamelCase = [0 for i in range(0 , 4 * size )] # create array to store lazy update _UpperCamelCase = [0 for i in range(0 , 4 * size )] _UpperCamelCase = [0 for i in range(0 , 4 * size )] # flag for lazy update def UpperCamelCase_ ( self : str , _A : int ): return idx * 2 def UpperCamelCase_ ( self : Any , _A : int ): return idx * 2 + 1 def UpperCamelCase_ ( self : Union[str, Any] , _A : int , _A : int , _A : int , _A : list[int] ): if left_element == right_element: _UpperCamelCase = a[left_element - 1] else: _UpperCamelCase = (left_element + right_element) // 2 self.build(self.left(_A ) , _A , _A , _A ) self.build(self.right(_A ) , mid + 1 , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) def UpperCamelCase_ ( self : Tuple , _A : int , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return True if left_element >= a and right_element <= b: _UpperCamelCase = val if left_element != right_element: _UpperCamelCase = val _UpperCamelCase = val _UpperCamelCase = True _UpperCamelCase = True return True _UpperCamelCase = (left_element + right_element) // 2 self.update(self.left(_A ) , _A , _A , _A , _A , _A ) self.update(self.right(_A ) , mid + 1 , _A , _A , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) return True def UpperCamelCase_ ( self : Any , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return -math.inf if left_element >= a and right_element <= b: return self.segment_tree[idx] _UpperCamelCase = (left_element + right_element) // 2 _UpperCamelCase = self.query(self.left(_A ) , _A , _A , _A , _A ) _UpperCamelCase = self.query(self.right(_A ) , mid + 1 , _A , _A , _A ) return max(_A , _A ) def __str__( self : Tuple ): return str([self.query(1 , 1 , self.size , _A , _A ) for i in range(1 , self.size + 1 )] ) if __name__ == "__main__": _lowerCAmelCase = [1, 2, -4, 7, 3, -5, 6, 11, -20, 9, 14, 15, 5, 2, -8] _lowerCAmelCase = 15 _lowerCAmelCase = SegmentTree(size) segt.build(1, 1, size, A) print(segt.query(1, 1, size, 4, 6)) print(segt.query(1, 1, size, 7, 11)) print(segt.query(1, 1, size, 7, 12)) segt.update(1, 1, size, 1, 3, 111) print(segt.query(1, 1, size, 1, 15)) segt.update(1, 1, size, 7, 8, 235) print(segt)
71
import torch from diffusers import DPMSolverSDEScheduler from diffusers.utils import torch_device from diffusers.utils.testing_utils import require_torchsde from .test_schedulers import SchedulerCommonTest @require_torchsde class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = (DPMSolverSDEScheduler,) UpperCAmelCase = 10 def UpperCamelCase_ ( self : Tuple , **_A : Union[str, Any] ): _UpperCamelCase = { '''num_train_timesteps''': 1100, '''beta_start''': 0.0001, '''beta_end''': 0.02, '''beta_schedule''': '''linear''', '''noise_sampler_seed''': 0, } config.update(**_A ) return config def UpperCamelCase_ ( self : List[Any] ): for timesteps in [10, 50, 100, 1000]: self.check_over_configs(num_train_timesteps=_A ) def UpperCamelCase_ ( self : List[Any] ): for beta_start, beta_end in zip([0.0_0001, 0.0001, 0.001] , [0.0002, 0.002, 0.02] ): self.check_over_configs(beta_start=_A , beta_end=_A ) def UpperCamelCase_ ( self : List[str] ): for schedule in ["linear", "scaled_linear"]: self.check_over_configs(beta_schedule=_A ) def UpperCamelCase_ ( self : Union[str, Any] ): for prediction_type in ["epsilon", "v_prediction"]: self.check_over_configs(prediction_type=_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for i, t in enumerate(scheduler.timesteps ): _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 167.47_8210_4492_1875 ) < 1e-2 assert abs(result_mean.item() - 0.2178_7059_6456_5277 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 171.59_3521_1181_6406 ) < 1e-2 assert abs(result_mean.item() - 0.2_2342_9068_9229_9652 ) < 1e-3 else: assert abs(result_sum.item() - 162.52_3834_2285_1562 ) < 1e-2 assert abs(result_mean.item() - 0.211_6195_7085_1326 ) < 1e-3 def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config(prediction_type='''v_prediction''' ) _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for i, t in enumerate(scheduler.timesteps ): _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 124.77_1492_0043_9453 ) < 1e-2 assert abs(result_mean.item() - 0.1_6226_2890_1481_6284 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 128.1_6633_6059_5703 ) < 1e-2 assert abs(result_mean.item() - 0.1_6688_3260_0116_7297 ) < 1e-3 else: assert abs(result_sum.item() - 119.8_4875_4882_8125 ) < 1e-2 assert abs(result_mean.item() - 0.1560_5306_6253_6621 ) < 1e-3 def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps , device=_A ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter.to(_A ) * scheduler.init_noise_sigma for t in scheduler.timesteps: _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 167.46_9573_9746_0938 ) < 1e-2 assert abs(result_mean.item() - 0.2_1805_9346_0798_2635 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 171.59_3536_3769_5312 ) < 1e-2 assert abs(result_mean.item() - 0.2_2342_9083_8241_5771 ) < 1e-3 else: assert abs(result_sum.item() - 162.52_3834_2285_1562 ) < 1e-2 assert abs(result_mean.item() - 0.211_6195_7085_1326 ) < 1e-3 def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A , use_karras_sigmas=_A ) scheduler.set_timesteps(self.num_inference_steps , device=_A ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter.to(_A ) * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for t in scheduler.timesteps: _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 176.66_9741_3574_2188 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 177.63_6535_6445_3125 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2 else: assert abs(result_sum.item() - 170.3_1352_2338_8672 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2
71
1
from __future__ import annotations def _snake_case ( __snake_case , __snake_case = None , __snake_case = None ): if start is None: _UpperCamelCase = 0 if end is None: _UpperCamelCase = len(__snake_case ) - 1 if start >= end: return _UpperCamelCase = (start + end) // 2 slowsort(__snake_case , __snake_case , __snake_case ) slowsort(__snake_case , mid + 1 , __snake_case ) if sequence[end] < sequence[mid]: _UpperCamelCase , _UpperCamelCase = sequence[mid], sequence[end] slowsort(__snake_case , __snake_case , end - 1 ) if __name__ == "__main__": from doctest import testmod testmod()
71
import unittest from typing import Tuple import torch from diffusers.utils import floats_tensor, randn_tensor, torch_all_close, torch_device from diffusers.utils.testing_utils import require_torch @require_torch class lowerCAmelCase_ : @property def UpperCamelCase_ ( self : Optional[int] ): return self.get_dummy_input() @property def UpperCamelCase_ ( self : Dict ): if self.block_type == "down": return (4, 32, 16, 16) elif self.block_type == "mid": return (4, 32, 32, 32) elif self.block_type == "up": return (4, 32, 64, 64) raise ValueError(F"""'{self.block_type}' is not a supported block_type. Set it to 'up', 'mid', or 'down'.""" ) def UpperCamelCase_ ( self : Union[str, Any] , _A : List[str]=True , _A : Any=False , _A : Union[str, Any]=False , _A : int=False , ): _UpperCamelCase = 4 _UpperCamelCase = 32 _UpperCamelCase = (32, 32) _UpperCamelCase = torch.manual_seed(0 ) _UpperCamelCase = torch.device(_A ) _UpperCamelCase = (batch_size, num_channels) + sizes _UpperCamelCase = randn_tensor(_A , generator=_A , device=_A ) _UpperCamelCase = {'''hidden_states''': hidden_states} if include_temb: _UpperCamelCase = 128 _UpperCamelCase = randn_tensor((batch_size, temb_channels) , generator=_A , device=_A ) if include_res_hidden_states_tuple: _UpperCamelCase = torch.manual_seed(1 ) _UpperCamelCase = (randn_tensor(_A , generator=_A , device=_A ),) if include_encoder_hidden_states: _UpperCamelCase = floats_tensor((batch_size, 32, 32) ).to(_A ) if include_skip_sample: _UpperCamelCase = randn_tensor(((batch_size, 3) + sizes) , generator=_A , device=_A ) return dummy_input def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = { '''in_channels''': 32, '''out_channels''': 32, '''temb_channels''': 128, } if self.block_type == "up": _UpperCamelCase = 32 if self.block_type == "mid": init_dict.pop('''out_channels''' ) _UpperCamelCase = self.dummy_input return init_dict, inputs_dict def UpperCamelCase_ ( self : Tuple , _A : Union[str, Any] ): _UpperCamelCase , _UpperCamelCase = self.prepare_init_args_and_inputs_for_common() _UpperCamelCase = self.block_class(**_A ) unet_block.to(_A ) unet_block.eval() with torch.no_grad(): _UpperCamelCase = unet_block(**_A ) if isinstance(_A , _A ): _UpperCamelCase = output[0] self.assertEqual(output.shape , self.output_shape ) _UpperCamelCase = output[0, -1, -3:, -3:] _UpperCamelCase = torch.tensor(_A ).to(_A ) assert torch_all_close(output_slice.flatten() , _A , atol=5e-3 ) @unittest.skipIf(torch_device == '''mps''' , '''Training is not supported in mps''' ) def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase , _UpperCamelCase = self.prepare_init_args_and_inputs_for_common() _UpperCamelCase = self.block_class(**_A ) model.to(_A ) model.train() _UpperCamelCase = model(**_A ) if isinstance(_A , _A ): _UpperCamelCase = output[0] _UpperCamelCase = torch.device(_A ) _UpperCamelCase = randn_tensor(output.shape , device=_A ) _UpperCamelCase = torch.nn.functional.mse_loss(_A , _A ) loss.backward()
71
1
import tempfile import unittest from transformers import SPIECE_UNDERLINE, BatchEncoding, PLBartTokenizer, is_torch_available from transformers.testing_utils import ( get_tests_dir, nested_simplify, require_sentencepiece, require_tokenizers, require_torch, ) from ...test_tokenization_common import TokenizerTesterMixin _lowerCAmelCase = get_tests_dir("fixtures/test_sentencepiece.model") if is_torch_available(): from transformers.models.plbart.modeling_plbart import shift_tokens_right _lowerCAmelCase = 50_003 _lowerCAmelCase = 50_002 @require_sentencepiece @require_tokenizers class lowerCAmelCase_ ( __lowercase, unittest.TestCase ): UpperCAmelCase = PLBartTokenizer UpperCAmelCase = None UpperCAmelCase = False def UpperCamelCase_ ( self : Any ): super().setUp() # We have a SentencePiece fixture for testing _UpperCamelCase = PLBartTokenizer(_A , language_codes='''base''' , keep_accents=_A ) tokenizer.save_pretrained(self.tmpdirname ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = PLBartTokenizer(_A , language_codes='''base''' , keep_accents=_A ) _UpperCamelCase = tokenizer.tokenize('''This is a test''' ) self.assertListEqual(_A , ['''▁This''', '''▁is''', '''▁a''', '''▁t''', '''est'''] ) self.assertListEqual( tokenizer.convert_tokens_to_ids(_A ) , [value + tokenizer.fairseq_offset for value in [285, 46, 10, 170, 382]] , ) _UpperCamelCase = tokenizer.tokenize('''I was born in 92000, and this is falsé.''' ) self.assertListEqual( _A , [ SPIECE_UNDERLINE + '''I''', SPIECE_UNDERLINE + '''was''', SPIECE_UNDERLINE + '''b''', '''or''', '''n''', SPIECE_UNDERLINE + '''in''', SPIECE_UNDERLINE + '''''', '''9''', '''2''', '''0''', '''0''', '''0''', ''',''', SPIECE_UNDERLINE + '''and''', SPIECE_UNDERLINE + '''this''', SPIECE_UNDERLINE + '''is''', SPIECE_UNDERLINE + '''f''', '''al''', '''s''', '''é''', '''.''', ] , ) _UpperCamelCase = tokenizer.convert_tokens_to_ids(_A ) self.assertListEqual( _A , [ value + tokenizer.fairseq_offset for value in [8, 21, 84, 55, 24, 19, 7, 2, 602, 347, 347, 347, 3, 12, 66, 46, 72, 80, 6, 2, 4] ] , ) _UpperCamelCase = tokenizer.convert_ids_to_tokens(_A ) self.assertListEqual( _A , [ SPIECE_UNDERLINE + '''I''', SPIECE_UNDERLINE + '''was''', SPIECE_UNDERLINE + '''b''', '''or''', '''n''', SPIECE_UNDERLINE + '''in''', SPIECE_UNDERLINE + '''''', '''<unk>''', '''2''', '''0''', '''0''', '''0''', ''',''', SPIECE_UNDERLINE + '''and''', SPIECE_UNDERLINE + '''this''', SPIECE_UNDERLINE + '''is''', SPIECE_UNDERLINE + '''f''', '''al''', '''s''', '''<unk>''', '''.''', ] , ) _UpperCamelCase = tokenizer.vocab_size _UpperCamelCase = [tokenizer.convert_ids_to_tokens(_A ) for x in range(end - 4 , _A )] self.assertListEqual(_A , ['''__java__''', '''__python__''', '''__en_XX__''', '''<mask>'''] ) _UpperCamelCase = '''java.lang.Exception, python.lang.Exception, javascript, php, ruby, go''' _UpperCamelCase = tokenizer(_A ).input_ids self.assertEqual( tokenizer.decode(_A , skip_special_tokens=_A , clean_up_tokenization_spaces=_A ) , _A , ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = PLBartTokenizer(_A , language_codes='''multi''' , keep_accents=_A ) _UpperCamelCase = tokenizer.tokenize('''This is a test''' ) self.assertListEqual(_A , ['''▁This''', '''▁is''', '''▁a''', '''▁t''', '''est'''] ) self.assertListEqual( tokenizer.convert_tokens_to_ids(_A ) , [value + tokenizer.fairseq_offset for value in [285, 46, 10, 170, 382]] , ) _UpperCamelCase = tokenizer.tokenize('''I was born in 92000, and this is falsé.''' ) self.assertListEqual( _A , [ SPIECE_UNDERLINE + '''I''', SPIECE_UNDERLINE + '''was''', SPIECE_UNDERLINE + '''b''', '''or''', '''n''', SPIECE_UNDERLINE + '''in''', SPIECE_UNDERLINE + '''''', '''9''', '''2''', '''0''', '''0''', '''0''', ''',''', SPIECE_UNDERLINE + '''and''', SPIECE_UNDERLINE + '''this''', SPIECE_UNDERLINE + '''is''', SPIECE_UNDERLINE + '''f''', '''al''', '''s''', '''é''', '''.''', ] , ) _UpperCamelCase = tokenizer.convert_tokens_to_ids(_A ) self.assertListEqual( _A , [ value + tokenizer.fairseq_offset for value in [8, 21, 84, 55, 24, 19, 7, 2, 602, 347, 347, 347, 3, 12, 66, 46, 72, 80, 6, 2, 4] ] , ) _UpperCamelCase = tokenizer.convert_ids_to_tokens(_A ) self.assertListEqual( _A , [ SPIECE_UNDERLINE + '''I''', SPIECE_UNDERLINE + '''was''', SPIECE_UNDERLINE + '''b''', '''or''', '''n''', SPIECE_UNDERLINE + '''in''', SPIECE_UNDERLINE + '''''', '''<unk>''', '''2''', '''0''', '''0''', '''0''', ''',''', SPIECE_UNDERLINE + '''and''', SPIECE_UNDERLINE + '''this''', SPIECE_UNDERLINE + '''is''', SPIECE_UNDERLINE + '''f''', '''al''', '''s''', '''<unk>''', '''.''', ] , ) _UpperCamelCase = tokenizer.vocab_size _UpperCamelCase = [tokenizer.convert_ids_to_tokens(_A ) for x in range(end - 7 , _A )] self.assertListEqual( _A , ['''__java__''', '''__python__''', '''__en_XX__''', '''__javascript__''', '''__php__''', '''__ruby__''', '''__go__'''] ) _UpperCamelCase = '''java.lang.Exception, python.lang.Exception, javascript, php, ruby, go''' _UpperCamelCase = tokenizer(_A ).input_ids self.assertEqual( tokenizer.decode(_A , skip_special_tokens=_A , clean_up_tokenization_spaces=_A ) , _A , ) @require_torch @require_sentencepiece @require_tokenizers class lowerCAmelCase_ ( unittest.TestCase ): UpperCAmelCase = "uclanlp/plbart-python-en_XX" UpperCAmelCase = [ "def maximum(a,b,c):NEW_LINE_INDENTreturn max([a,b,c])", "def sum(a,b,c):NEW_LINE_INDENTreturn sum([a,b,c])", ] UpperCAmelCase = [ "Returns the maximum value of a b c.", "Sums the values of a b c.", ] UpperCAmelCase = [ 134, 5452, 33460, 33441, 33463, 33465, 33463, 33449, 988, 20, 33456, 19, 33456, 771, 39, 4258, 889, 3318, 33441, 33463, 33465, 33463, 33449, 2471, 2, PYTHON_CODE, ] @classmethod def UpperCamelCase_ ( cls : List[str] ): _UpperCamelCase = PLBartTokenizer.from_pretrained( cls.checkpoint_name , language_codes='''base''' , src_lang='''python''' , tgt_lang='''en_XX''' ) _UpperCamelCase = 1 return cls def UpperCamelCase_ ( self : Optional[Any] ): self.assertEqual(self.tokenizer.fairseq_tokens_to_ids['''__java__'''] , 5_0001 ) self.assertEqual(self.tokenizer.fairseq_tokens_to_ids['''__python__'''] , 5_0002 ) self.assertEqual(self.tokenizer.fairseq_tokens_to_ids['''__en_XX__'''] , 5_0003 ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.tokenizer.batch_encode_plus(self.src_text ).input_ids[0] self.assertListEqual(self.expected_src_tokens , _A ) def UpperCamelCase_ ( self : Union[str, Any] ): self.assertIn(_A , self.tokenizer.all_special_ids ) _UpperCamelCase = [EN_CODE, 9037, 3_3442, 57, 752, 153, 14, 56, 18, 9, 2] _UpperCamelCase = self.tokenizer.decode(_A , skip_special_tokens=_A ) _UpperCamelCase = self.tokenizer.decode(generated_ids[1:] , skip_special_tokens=_A ) self.assertEqual(_A , _A ) self.assertNotIn(self.tokenizer.eos_token , _A ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = ['''def sum(a,b,c):NEW_LINE_INDENTreturn sum([a,b,c])''' * 20] self.assertIsInstance(src_text[0] , _A ) _UpperCamelCase = 10 _UpperCamelCase = self.tokenizer(_A , max_length=_A , truncation=_A ).input_ids[0] self.assertEqual(ids[-2] , 2 ) self.assertEqual(ids[-1] , _A ) self.assertEqual(len(_A ) , _A ) def UpperCamelCase_ ( self : List[Any] ): self.assertListEqual(self.tokenizer.convert_tokens_to_ids(['''<mask>''', '''__java__'''] ) , [5_0004, 5_0001] ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = tempfile.mkdtemp() _UpperCamelCase = self.tokenizer.fairseq_tokens_to_ids self.tokenizer.save_pretrained(_A ) _UpperCamelCase = PLBartTokenizer.from_pretrained(_A ) self.assertDictEqual(new_tok.fairseq_tokens_to_ids , _A ) @require_torch def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.tokenizer(self.src_text , text_target=self.tgt_text , padding=_A , return_tensors='''pt''' ) _UpperCamelCase = shift_tokens_right(batch['''labels'''] , self.tokenizer.pad_token_id ) # fairseq batch: https://gist.github.com/sshleifer/cba08bc2109361a74ac3760a7e30e4f4 self.assertEqual(batch.input_ids[1][-2:].tolist() , [2, PYTHON_CODE] ) self.assertEqual(batch.decoder_input_ids[1][0] , _A ) self.assertEqual(batch.decoder_input_ids[1][-1] , 2 ) self.assertEqual(batch.labels[1][-2:].tolist() , [2, EN_CODE] ) @require_torch def UpperCamelCase_ ( self : Any ): _UpperCamelCase = self.tokenizer( self.src_text , text_target=self.tgt_text , padding=_A , truncation=_A , max_length=len(self.expected_src_tokens ) , return_tensors='''pt''' , ) _UpperCamelCase = shift_tokens_right(batch['''labels'''] , self.tokenizer.pad_token_id ) self.assertIsInstance(_A , _A ) self.assertEqual((2, 26) , batch.input_ids.shape ) self.assertEqual((2, 26) , batch.attention_mask.shape ) _UpperCamelCase = batch.input_ids.tolist()[0] self.assertListEqual(self.expected_src_tokens , _A ) self.assertEqual(2 , batch.decoder_input_ids[0, -1] ) # EOS # Test that special tokens are reset self.assertEqual(self.tokenizer.prefix_tokens , [] ) self.assertEqual(self.tokenizer.suffix_tokens , [self.tokenizer.eos_token_id, PYTHON_CODE] ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.tokenizer(self.src_text , padding=_A , truncation=_A , max_length=3 , return_tensors='''pt''' ) _UpperCamelCase = self.tokenizer( text_target=self.tgt_text , padding=_A , truncation=_A , max_length=10 , return_tensors='''pt''' ) _UpperCamelCase = targets['''input_ids'''] _UpperCamelCase = shift_tokens_right(_A , self.tokenizer.pad_token_id ) self.assertEqual(batch.input_ids.shape[1] , 3 ) self.assertEqual(batch.decoder_input_ids.shape[1] , 10 ) @require_torch def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.tokenizer._build_translation_inputs( '''A test''' , return_tensors='''pt''' , src_lang='''en_XX''' , tgt_lang='''java''' ) self.assertEqual( nested_simplify(_A ) , { # A, test, EOS, en_XX '''input_ids''': [[150, 242, 2, 5_0003]], '''attention_mask''': [[1, 1, 1, 1]], # java '''forced_bos_token_id''': 5_0001, } , )
71
def _snake_case ( __snake_case ): if not isinstance(__snake_case , __snake_case ): raise TypeError('''Input value must be an \'int\' type''' ) _UpperCamelCase = 0 while number: position += 1 number >>= 1 return position if __name__ == "__main__": import doctest doctest.testmod()
71
1
import argparse from pathlib import Path from transformers import AutoConfig, AutoTokenizer, RagConfig, RagSequenceForGeneration, RagTokenForGeneration def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case = None , __snake_case = None , __snake_case = None , ): if config_name_or_path is None: _UpperCamelCase = '''facebook/rag-token-base''' if model_type == '''rag_token''' else '''facebook/rag-sequence-base''' if generator_tokenizer_name_or_path is None: _UpperCamelCase = generator_name_or_path if question_encoder_tokenizer_name_or_path is None: _UpperCamelCase = question_encoder_name_or_path _UpperCamelCase = RagTokenForGeneration if model_type == '''rag_token''' else RagSequenceForGeneration # Save model. _UpperCamelCase = RagConfig.from_pretrained(__snake_case ) _UpperCamelCase = AutoConfig.from_pretrained(__snake_case ) _UpperCamelCase = AutoConfig.from_pretrained(__snake_case ) _UpperCamelCase = gen_config _UpperCamelCase = question_encoder_config _UpperCamelCase = model_class.from_pretrained_question_encoder_generator( __snake_case , __snake_case , config=__snake_case ) rag_model.save_pretrained(__snake_case ) # Sanity check. model_class.from_pretrained(__snake_case ) # Save tokenizers. _UpperCamelCase = AutoTokenizer.from_pretrained(__snake_case ) gen_tokenizer.save_pretrained(dest_dir / '''generator_tokenizer/''' ) _UpperCamelCase = AutoTokenizer.from_pretrained(__snake_case ) question_encoder_tokenizer.save_pretrained(dest_dir / '''question_encoder_tokenizer/''' ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() parser.add_argument( "--model_type", choices=["rag_sequence", "rag_token"], required=True, type=str, help="RAG model type: rag_sequence, rag_token", ) parser.add_argument("--dest", type=str, required=True, help="Path to the output checkpoint directory.") parser.add_argument("--generator_name_or_path", type=str, required=True, help="Generator model identifier") parser.add_argument( "--question_encoder_name_or_path", type=str, required=True, help="Question encoder model identifier" ) parser.add_argument( "--generator_tokenizer_name_or_path", type=str, help="Generator tokenizer identifier, if not specified, resolves to ``generator_name_or_path``", ) parser.add_argument( "--question_encoder_tokenizer_name_or_path", type=str, help="Question encoder tokenizer identifier, if not specified, resolves to ``question_encoder_name_or_path``", ) parser.add_argument( "--config_name_or_path", type=str, help=( "Identifier of the model config to use, if not provided, resolves to a base config for a given" " ``model_type``" ), ) _lowerCAmelCase = parser.parse_args() _lowerCAmelCase = Path(args.dest) dest_dir.mkdir(exist_ok=True) consolidate( args.model_type, args.generator_name_or_path, args.question_encoder_name_or_path, dest_dir, args.config_name_or_path, args.generator_tokenizer_name_or_path, args.question_encoder_tokenizer_name_or_path, )
71
import argparse import json import os import fairseq import torch from torch import nn from transformers import ( SpeechaTextaConfig, SpeechaTextaForCausalLM, SpeechaTextaTokenizer, SpeechEncoderDecoderConfig, SpeechEncoderDecoderModel, WavaVecaConfig, WavaVecaFeatureExtractor, WavaVecaModel, logging, ) logging.set_verbosity_info() _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "post_extract_proj": "feature_projection.projection", "encoder.pos_conv.0": "encoder.pos_conv_embed.conv", "self_attn.k_proj": "encoder.layers.*.attention.k_proj", "self_attn.v_proj": "encoder.layers.*.attention.v_proj", "self_attn.q_proj": "encoder.layers.*.attention.q_proj", "self_attn.out_proj": "encoder.layers.*.attention.out_proj", "self_attn_layer_norm": "encoder.layers.*.layer_norm", "fc1": "encoder.layers.*.feed_forward.intermediate_dense", "fc2": "encoder.layers.*.feed_forward.output_dense", "final_layer_norm": "encoder.layers.*.final_layer_norm", "encoder.layer_norm": "encoder.layer_norm", "w2v_model.layer_norm": "feature_projection.layer_norm", "quantizer.weight_proj": "quantizer.weight_proj", "quantizer.vars": "quantizer.codevectors", "project_q": "project_q", "final_proj": "project_hid", "w2v_encoder.proj": "lm_head", "mask_emb": "masked_spec_embed", } _lowerCAmelCase = [ "lm_head", "quantizer.weight_proj", "quantizer.codevectors", "project_q", "project_hid", ] def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): for attribute in key.split('''.''' ): _UpperCamelCase = getattr(__snake_case , __snake_case ) if weight_type is not None: _UpperCamelCase = getattr(__snake_case , __snake_case ).shape else: _UpperCamelCase = hf_pointer.shape assert hf_shape == value.shape, ( f"""Shape of hf {key + "." + weight_type if weight_type is not None else ""} is {hf_shape}, but should be""" f""" {value.shape} for {full_name}""" ) if weight_type == "weight": _UpperCamelCase = value elif weight_type == "weight_g": _UpperCamelCase = value elif weight_type == "weight_v": _UpperCamelCase = value elif weight_type == "bias": _UpperCamelCase = value else: _UpperCamelCase = value logger.info(f"""{key + "." + weight_type if weight_type is not None else ""} was initialized from {full_name}.""" ) def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = [] _UpperCamelCase = fairseq_model.state_dict() _UpperCamelCase = hf_model.feature_extractor # if encoder has different dim to decoder -> use proj_weight _UpperCamelCase = None for name, value in fairseq_dict.items(): _UpperCamelCase = False if "conv_layers" in name: load_conv_layer( __snake_case , __snake_case , __snake_case , __snake_case , hf_model.config.feat_extract_norm == '''group''' , ) _UpperCamelCase = True elif name.split('''.''' )[0] == "proj": _UpperCamelCase = fairseq_model.proj _UpperCamelCase = True else: for key, mapped_key in MAPPING.items(): if key in name or key.split('''w2v_model.''' )[-1] == name.split('''.''' )[0]: _UpperCamelCase = True if "*" in mapped_key: _UpperCamelCase = name.split(__snake_case )[0].split('''.''' )[-2] _UpperCamelCase = mapped_key.replace('''*''' , __snake_case ) if "weight_g" in name: _UpperCamelCase = '''weight_g''' elif "weight_v" in name: _UpperCamelCase = '''weight_v''' elif "bias" in name: _UpperCamelCase = '''bias''' elif "weight" in name: _UpperCamelCase = '''weight''' else: _UpperCamelCase = None set_recursively(__snake_case , __snake_case , __snake_case , __snake_case , __snake_case ) continue if not is_used: unused_weights.append(__snake_case ) logger.warning(f"""Unused weights: {unused_weights}""" ) return proj_weight def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = full_name.split('''conv_layers.''' )[-1] _UpperCamelCase = name.split('''.''' ) _UpperCamelCase = int(items[0] ) _UpperCamelCase = int(items[1] ) if type_id == 0: if "bias" in name: assert value.shape == feature_extractor.conv_layers[layer_id].conv.bias.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor.conv_layers[layer_id].conv.bias.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract conv layer {layer_id} was initialized from {full_name}.""" ) elif "weight" in name: assert value.shape == feature_extractor.conv_layers[layer_id].conv.weight.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor.conv_layers[layer_id].conv.weight.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract conv layer {layer_id} was initialized from {full_name}.""" ) elif (type_id == 2 and not use_group_norm) or (type_id == 2 and layer_id == 0 and use_group_norm): if "bias" in name: assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.bias.data.shape, ( f"""{full_name} has size {value.shape}, but {feature_extractor[layer_id].layer_norm.bias.data.shape} was""" " found." ) _UpperCamelCase = value logger.info(f"""Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.""" ) elif "weight" in name: assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.weight.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor[layer_id].layer_norm.weight.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.""" ) else: unused_weights.append(__snake_case ) def _snake_case ( __snake_case ): _UpperCamelCase , _UpperCamelCase = emb.weight.shape _UpperCamelCase = nn.Linear(__snake_case , __snake_case , bias=__snake_case ) _UpperCamelCase = emb.weight.data return lin_layer def _snake_case ( __snake_case ): with open(__snake_case , '''r''' , encoding='''utf-8''' ) as f: _UpperCamelCase = f.readlines() _UpperCamelCase = [line.split(''' ''' )[0] for line in lines] _UpperCamelCase = len(__snake_case ) _UpperCamelCase = { '''<s>''': 0, '''<pad>''': 1, '''</s>''': 2, '''<unk>''': 3, } vocab_dict.update(dict(zip(__snake_case , range(4 , num_words + 4 ) ) ) ) return vocab_dict @torch.no_grad() def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , ): _UpperCamelCase = WavaVecaConfig.from_pretrained(__snake_case ) _UpperCamelCase = SpeechaTextaConfig.from_pretrained( __snake_case , vocab_size=__snake_case , decoder_layers=__snake_case , do_stable_layer_norm=__snake_case ) _UpperCamelCase = WavaVecaFeatureExtractor( feature_size=1 , sampling_rate=16000 , padding_value=0 , do_normalize=__snake_case , return_attention_mask=__snake_case , ) _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = fairseq.checkpoint_utils.load_model_ensemble_and_task( [checkpoint_path] , arg_overrides={'''data''': '''/'''.join(dict_path.split('''/''' )[:-1] )} ) _UpperCamelCase = model[0].eval() # set weights for wav2vec2 encoder _UpperCamelCase = WavaVecaModel(__snake_case ) _UpperCamelCase = recursively_load_weights_wavaveca(model.encoder , __snake_case ) _UpperCamelCase = SpeechaTextaForCausalLM(__snake_case ) _UpperCamelCase , _UpperCamelCase = hf_decoder.model.decoder.load_state_dict(model.decoder.state_dict() , strict=__snake_case ) # set output linear layer unexpected_keys.remove('''embed_out''' ) _UpperCamelCase = nn.Parameter(model.decoder.embed_out.detach() ) # layer norm is init to identity matrix so leaving it is fine logger.warning(f"""The following keys are missing when loading the decoder weights: {missing_keys}""" ) logger.warning(f"""The following keys are unexpected when loading the decoder weights: {unexpected_keys}""" ) _UpperCamelCase = SpeechEncoderDecoderModel(encoder=__snake_case , decoder=__snake_case ) _UpperCamelCase = False # add projection layer _UpperCamelCase = nn.Parameter(projection_layer.weight ) _UpperCamelCase = nn.Parameter(projection_layer.bias ) _UpperCamelCase = create_vocab_dict(__snake_case ) with open(os.path.join(__snake_case , '''vocab.json''' ) , '''w''' ) as fp: json.dump(__snake_case , __snake_case ) _UpperCamelCase = SpeechaTextaTokenizer(os.path.join(__snake_case , '''vocab.json''' ) ) tokenizer.save_pretrained(__snake_case ) _UpperCamelCase = hf_wavavec.config.to_dict() _UpperCamelCase = tokenizer.pad_token_id _UpperCamelCase = tokenizer.bos_token_id _UpperCamelCase = tokenizer.eos_token_id _UpperCamelCase = '''speech_to_text_2''' _UpperCamelCase = '''wav2vec2''' _UpperCamelCase = SpeechEncoderDecoderConfig.from_dict(__snake_case ) hf_wavavec.save_pretrained(__snake_case ) feature_extractor.save_pretrained(__snake_case ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() parser.add_argument("--pytorch_dump_folder_path", default=None, type=str, help="Path to the output PyTorch model.") parser.add_argument("--checkpoint_path", default=None, type=str, help="Path to fairseq checkpoint") parser.add_argument("--dict_path", default=None, type=str, help="Path to dict of fine-tuned model") parser.add_argument( "--encoder_config_path", default="facebook/wav2vec2-large-lv60", type=str, help="Path to hf encoder wav2vec2 checkpoint config", ) parser.add_argument( "--decoder_config_path", default="facebook/s2t-small-mustc-en-fr-st", type=str, help="Path to hf decoder s2t checkpoint config", ) parser.add_argument("--vocab_size", default=10_224, type=int, help="Vocab size of decoder") parser.add_argument("--num_decoder_layers", default=7, type=int, help="Number of decoder layers") _lowerCAmelCase = parser.parse_args() convert_wavaveca_checkpoint( args.checkpoint_path, args.pytorch_dump_folder_path, args.dict_path, encoder_config_path=args.encoder_config_path, decoder_config_path=args.decoder_config_path, vocab_size=args.vocab_size, num_decoder_layers=args.num_decoder_layers, )
71
1
from itertools import permutations def _snake_case ( __snake_case ): if num[3] % 2 != 0: return False if (num[2] + num[3] + num[4]) % 3 != 0: return False if num[5] % 5 != 0: return False _UpperCamelCase = [7, 11, 13, 17] for i, test in enumerate(__snake_case ): if (num[i + 4] * 100 + num[i + 5] * 10 + num[i + 6]) % test != 0: return False return True def _snake_case ( __snake_case = 10 ): return sum( int(''''''.join(map(__snake_case , __snake_case ) ) ) for num in permutations(range(__snake_case ) ) if is_substring_divisible(__snake_case ) ) if __name__ == "__main__": print(f'{solution() = }')
71
from __future__ import annotations import unittest from transformers import DebertaVaConfig, is_tf_available from transformers.testing_utils import require_tf, slow from ...test_configuration_common import ConfigTester from ...test_modeling_tf_common import TFModelTesterMixin, ids_tensor, random_attention_mask from ...test_pipeline_mixin import PipelineTesterMixin if is_tf_available(): import tensorflow as tf from transformers import ( TFDebertaVaForMaskedLM, TFDebertaVaForQuestionAnswering, TFDebertaVaForSequenceClassification, TFDebertaVaForTokenClassification, TFDebertaVaModel, ) class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : Optional[Any] , _A : List[str]=13 , _A : Union[str, Any]=7 , _A : int=True , _A : Optional[int]=True , _A : Optional[int]=True , _A : Union[str, Any]=True , _A : Optional[int]=99 , _A : Union[str, Any]=32 , _A : Dict=2 , _A : List[Any]=4 , _A : Optional[Any]=37 , _A : int="gelu" , _A : Optional[int]=0.1 , _A : str=0.1 , _A : List[str]=512 , _A : Optional[Any]=16 , _A : Optional[Any]=2 , _A : Optional[int]=0.02 , _A : str=False , _A : int=True , _A : Any="None" , _A : Dict=3 , _A : List[Any]=4 , _A : Optional[Any]=None , ): _UpperCamelCase = parent _UpperCamelCase = batch_size _UpperCamelCase = seq_length _UpperCamelCase = is_training _UpperCamelCase = use_input_mask _UpperCamelCase = use_token_type_ids _UpperCamelCase = use_labels _UpperCamelCase = vocab_size _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = hidden_dropout_prob _UpperCamelCase = attention_probs_dropout_prob _UpperCamelCase = max_position_embeddings _UpperCamelCase = type_vocab_size _UpperCamelCase = type_sequence_label_size _UpperCamelCase = initializer_range _UpperCamelCase = num_labels _UpperCamelCase = num_choices _UpperCamelCase = relative_attention _UpperCamelCase = position_biased_input _UpperCamelCase = pos_att_type _UpperCamelCase = scope def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size ) _UpperCamelCase = None if self.use_input_mask: _UpperCamelCase = random_attention_mask([self.batch_size, self.seq_length] ) _UpperCamelCase = None if self.use_token_type_ids: _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.type_vocab_size ) _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None if self.use_labels: _UpperCamelCase = ids_tensor([self.batch_size] , self.type_sequence_label_size ) _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.num_labels ) _UpperCamelCase = DebertaVaConfig( vocab_size=self.vocab_size , hidden_size=self.hidden_size , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , relative_attention=self.relative_attention , position_biased_input=self.position_biased_input , initializer_range=self.initializer_range , return_dict=_A , ) return config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels def UpperCamelCase_ ( self : Dict , _A : Tuple , _A : Tuple , _A : Union[str, Any] , _A : List[str] , _A : Optional[int] , _A : int , _A : Optional[Any] ): _UpperCamelCase = TFDebertaVaModel(config=_A ) _UpperCamelCase = {'''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids} _UpperCamelCase = [input_ids, input_mask] _UpperCamelCase = model(_A ) _UpperCamelCase = model(_A ) self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) ) def UpperCamelCase_ ( self : Dict , _A : Optional[int] , _A : Any , _A : Dict , _A : Union[str, Any] , _A : Union[str, Any] , _A : List[Any] , _A : List[str] ): _UpperCamelCase = TFDebertaVaForMaskedLM(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) ) def UpperCamelCase_ ( self : Dict , _A : Dict , _A : List[str] , _A : List[Any] , _A : List[Any] , _A : Optional[Any] , _A : Tuple , _A : int ): _UpperCamelCase = self.num_labels _UpperCamelCase = TFDebertaVaForSequenceClassification(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) ) def UpperCamelCase_ ( self : Tuple , _A : Dict , _A : Optional[int] , _A : Any , _A : List[Any] , _A : Dict , _A : Union[str, Any] , _A : List[str] ): _UpperCamelCase = self.num_labels _UpperCamelCase = TFDebertaVaForTokenClassification(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) ) def UpperCamelCase_ ( self : Dict , _A : Optional[Any] , _A : Optional[int] , _A : Any , _A : List[str] , _A : str , _A : Optional[int] , _A : str ): _UpperCamelCase = TFDebertaVaForQuestionAnswering(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.start_logits.shape , (self.batch_size, self.seq_length) ) self.parent.assertEqual(result.end_logits.shape , (self.batch_size, self.seq_length) ) def UpperCamelCase_ ( self : Any ): _UpperCamelCase = self.prepare_config_and_inputs() ( ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ) = config_and_inputs _UpperCamelCase = {'''input_ids''': input_ids, '''token_type_ids''': token_type_ids, '''attention_mask''': input_mask} return config, inputs_dict @require_tf class lowerCAmelCase_ ( __lowercase, __lowercase, unittest.TestCase ): UpperCAmelCase = ( ( TFDebertaVaModel, TFDebertaVaForMaskedLM, TFDebertaVaForQuestionAnswering, TFDebertaVaForSequenceClassification, TFDebertaVaForTokenClassification, ) if is_tf_available() else () ) UpperCAmelCase = ( { "feature-extraction": TFDebertaVaModel, "fill-mask": TFDebertaVaForMaskedLM, "question-answering": TFDebertaVaForQuestionAnswering, "text-classification": TFDebertaVaForSequenceClassification, "token-classification": TFDebertaVaForTokenClassification, "zero-shot": TFDebertaVaForSequenceClassification, } if is_tf_available() else {} ) UpperCAmelCase = False UpperCAmelCase = False def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = TFDebertaVaModelTester(self ) _UpperCamelCase = ConfigTester(self , config_class=_A , hidden_size=37 ) def UpperCamelCase_ ( self : Any ): self.config_tester.run_common_tests() def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model(*_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_masked_lm(*_A ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_question_answering(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_sequence_classification(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_token_classification(*_A ) @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = TFDebertaVaModel.from_pretrained('''kamalkraj/deberta-v2-xlarge''' ) self.assertIsNotNone(_A ) @require_tf class lowerCAmelCase_ ( unittest.TestCase ): @unittest.skip(reason='''Model not available yet''' ) def UpperCamelCase_ ( self : List[Any] ): pass @slow def UpperCamelCase_ ( self : int ): _UpperCamelCase = TFDebertaVaModel.from_pretrained('''kamalkraj/deberta-v2-xlarge''' ) _UpperCamelCase = tf.constant([[0, 3_1414, 232, 328, 740, 1140, 1_2695, 69, 4_6078, 1588, 2]] ) _UpperCamelCase = tf.constant([[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]] ) _UpperCamelCase = model(_A , attention_mask=_A )[0] _UpperCamelCase = tf.constant( [[[0.2356, 0.1948, 0.0369], [-0.1063, 0.3586, -0.5152], [-0.6399, -0.0259, -0.2525]]] ) tf.debugging.assert_near(output[:, 1:4, 1:4] , _A , atol=1e-4 )
71
1
import json import os import unittest from transformers.models.ctrl.tokenization_ctrl import VOCAB_FILES_NAMES, CTRLTokenizer from ...test_tokenization_common import TokenizerTesterMixin class lowerCAmelCase_ ( __lowercase, unittest.TestCase ): UpperCAmelCase = CTRLTokenizer UpperCAmelCase = False UpperCAmelCase = False def UpperCamelCase_ ( self : str ): super().setUp() # Adapted from Sennrich et al. 2015 and https://github.com/rsennrich/subword-nmt _UpperCamelCase = ['''adapt''', '''re@@''', '''a@@''', '''apt''', '''c@@''', '''t''', '''<unk>'''] _UpperCamelCase = dict(zip(_A , range(len(_A ) ) ) ) _UpperCamelCase = ['''#version: 0.2''', '''a p''', '''ap t</w>''', '''r e''', '''a d''', '''ad apt</w>''', ''''''] _UpperCamelCase = {'''unk_token''': '''<unk>'''} _UpperCamelCase = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES['''vocab_file'''] ) _UpperCamelCase = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES['''merges_file'''] ) with open(self.vocab_file , '''w''' , encoding='''utf-8''' ) as fp: fp.write(json.dumps(_A ) + '''\n''' ) with open(self.merges_file , '''w''' , encoding='''utf-8''' ) as fp: fp.write('''\n'''.join(_A ) ) def UpperCamelCase_ ( self : int , **_A : List[Any] ): kwargs.update(self.special_tokens_map ) return CTRLTokenizer.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : int , _A : int ): _UpperCamelCase = '''adapt react readapt apt''' _UpperCamelCase = '''adapt react readapt apt''' return input_text, output_text def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = CTRLTokenizer(self.vocab_file , self.merges_file , **self.special_tokens_map ) _UpperCamelCase = '''adapt react readapt apt''' _UpperCamelCase = '''adapt re@@ a@@ c@@ t re@@ adapt apt'''.split() _UpperCamelCase = tokenizer.tokenize(_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = tokens + [tokenizer.unk_token] _UpperCamelCase = [0, 1, 2, 4, 5, 1, 0, 3, 6] self.assertListEqual(tokenizer.convert_tokens_to_ids(_A ) , _A )
71
def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): # Return True if there is node that has not iterated. _UpperCamelCase = [False] * len(__snake_case ) _UpperCamelCase = [] queue.append(__snake_case ) _UpperCamelCase = True while queue: _UpperCamelCase = queue.pop(0 ) for ind in range(len(graph[u] ) ): if visited[ind] is False and graph[u][ind] > 0: queue.append(__snake_case ) _UpperCamelCase = True _UpperCamelCase = u return visited[t] def _snake_case ( __snake_case , __snake_case , __snake_case ): # This array is filled by BFS and to store path _UpperCamelCase = [-1] * (len(__snake_case )) _UpperCamelCase = 0 while bfs(__snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = float('''Inf''' ) _UpperCamelCase = sink while s != source: # Find the minimum value in select path _UpperCamelCase = min(__snake_case , graph[parent[s]][s] ) _UpperCamelCase = parent[s] max_flow += path_flow _UpperCamelCase = sink while v != source: _UpperCamelCase = parent[v] graph[u][v] -= path_flow graph[v][u] += path_flow _UpperCamelCase = parent[v] return max_flow _lowerCAmelCase = [ [0, 16, 13, 0, 0, 0], [0, 0, 10, 12, 0, 0], [0, 4, 0, 0, 14, 0], [0, 0, 9, 0, 0, 20], [0, 0, 0, 7, 0, 4], [0, 0, 0, 0, 0, 0], ] _lowerCAmelCase, _lowerCAmelCase = 0, 5 print(ford_fulkerson(graph, source, sink))
71
1
def _snake_case ( __snake_case , __snake_case ): return abs(__snake_case ) if a == 0 else greatest_common_divisor(b % a , __snake_case ) def _snake_case ( __snake_case , __snake_case ): while y: # --> when y=0 then loop will terminate and return x as final GCD. _UpperCamelCase , _UpperCamelCase = y, x % y return abs(__snake_case ) def _snake_case ( ): try: _UpperCamelCase = input('''Enter two integers separated by comma (,): ''' ).split(''',''' ) _UpperCamelCase = int(nums[0] ) _UpperCamelCase = int(nums[1] ) print( f"""greatest_common_divisor({num_a}, {num_a}) = """ f"""{greatest_common_divisor(__snake_case , __snake_case )}""" ) print(f"""By iterative gcd({num_a}, {num_a}) = {gcd_by_iterative(__snake_case , __snake_case )}""" ) except (IndexError, UnboundLocalError, ValueError): print('''Wrong input''' ) if __name__ == "__main__": main()
71
from typing import TYPE_CHECKING from ...utils import ( OptionalDependencyNotAvailable, _LazyModule, is_flax_available, is_tf_available, is_torch_available, ) _lowerCAmelCase = {"configuration_unispeech": ["UNISPEECH_PRETRAINED_CONFIG_ARCHIVE_MAP", "UniSpeechConfig"]} try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "UNISPEECH_PRETRAINED_MODEL_ARCHIVE_LIST", "UniSpeechForCTC", "UniSpeechForPreTraining", "UniSpeechForSequenceClassification", "UniSpeechModel", "UniSpeechPreTrainedModel", ] if TYPE_CHECKING: from .configuration_unispeech import UNISPEECH_PRETRAINED_CONFIG_ARCHIVE_MAP, UniSpeechConfig try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_unispeech import ( UNISPEECH_PRETRAINED_MODEL_ARCHIVE_LIST, UniSpeechForCTC, UniSpeechForPreTraining, UniSpeechForSequenceClassification, UniSpeechModel, UniSpeechPreTrainedModel, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
1
from typing import Mapping from ...configuration_utils import PretrainedConfig from ...onnx import OnnxSeqaSeqConfigWithPast from ...utils import logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "t5-small": "https://huggingface.co/t5-small/resolve/main/config.json", "t5-base": "https://huggingface.co/t5-base/resolve/main/config.json", "t5-large": "https://huggingface.co/t5-large/resolve/main/config.json", "t5-3b": "https://huggingface.co/t5-3b/resolve/main/config.json", "t5-11b": "https://huggingface.co/t5-11b/resolve/main/config.json", } class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "t5" UpperCAmelCase = ["past_key_values"] UpperCAmelCase = {"hidden_size": "d_model", "num_attention_heads": "num_heads", "num_hidden_layers": "num_layers"} def __init__( self : str , _A : List[str]=3_2128 , _A : Tuple=512 , _A : List[Any]=64 , _A : Tuple=2048 , _A : List[Any]=6 , _A : int=None , _A : List[str]=8 , _A : Optional[Any]=32 , _A : Tuple=128 , _A : Optional[int]=0.1 , _A : Dict=1e-6 , _A : List[Any]=1.0 , _A : Tuple="relu" , _A : Dict=True , _A : int=True , _A : Tuple=0 , _A : Optional[int]=1 , **_A : Tuple , ): _UpperCamelCase = vocab_size _UpperCamelCase = d_model _UpperCamelCase = d_kv _UpperCamelCase = d_ff _UpperCamelCase = num_layers _UpperCamelCase = ( num_decoder_layers if num_decoder_layers is not None else self.num_layers ) # default = symmetry _UpperCamelCase = num_heads _UpperCamelCase = relative_attention_num_buckets _UpperCamelCase = relative_attention_max_distance _UpperCamelCase = dropout_rate _UpperCamelCase = layer_norm_epsilon _UpperCamelCase = initializer_factor _UpperCamelCase = feed_forward_proj _UpperCamelCase = use_cache _UpperCamelCase = self.feed_forward_proj.split('''-''' ) _UpperCamelCase = act_info[-1] _UpperCamelCase = act_info[0] == '''gated''' if len(_A ) > 1 and act_info[0] != "gated" or len(_A ) > 2: raise ValueError( F"""`feed_forward_proj`: {feed_forward_proj} is not a valid activation function of the dense layer.""" '''Please make sure `feed_forward_proj` is of the format `gated-{ACT_FN}` or `{ACT_FN}`, e.g. ''' '''\'gated-gelu\' or \'relu\'''' ) # for backwards compatibility if feed_forward_proj == "gated-gelu": _UpperCamelCase = '''gelu_new''' super().__init__( pad_token_id=_A , eos_token_id=_A , is_encoder_decoder=_A , **_A , ) class lowerCAmelCase_ ( __lowercase ): @property def UpperCamelCase_ ( self : int ): _UpperCamelCase = { '''input_ids''': {0: '''batch''', 1: '''encoder_sequence'''}, '''attention_mask''': {0: '''batch''', 1: '''encoder_sequence'''}, } if self.use_past: _UpperCamelCase = '''past_encoder_sequence + sequence''' _UpperCamelCase = {0: '''batch'''} _UpperCamelCase = {0: '''batch''', 1: '''past_decoder_sequence + sequence'''} else: _UpperCamelCase = {0: '''batch''', 1: '''decoder_sequence'''} _UpperCamelCase = {0: '''batch''', 1: '''decoder_sequence'''} if self.use_past: self.fill_with_past_key_values_(_A , direction='''inputs''' ) return common_inputs @property def UpperCamelCase_ ( self : Any ): return 13
71
import json import os import shutil import tempfile import unittest import numpy as np from transformers import BertTokenizerFast from transformers.models.bert.tokenization_bert import VOCAB_FILES_NAMES, BertTokenizer from transformers.testing_utils import require_tokenizers, require_vision from transformers.utils import IMAGE_PROCESSOR_NAME, is_vision_available if is_vision_available(): from PIL import Image from transformers import VisionTextDualEncoderProcessor, ViTImageProcessor @require_tokenizers @require_vision class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = tempfile.mkdtemp() # fmt: off _UpperCamelCase = ['''[UNK]''', '''[CLS]''', '''[SEP]''', '''[PAD]''', '''[MASK]''', '''want''', '''##want''', '''##ed''', '''wa''', '''un''', '''runn''', '''##ing''', ''',''', '''low''', '''lowest'''] # fmt: on _UpperCamelCase = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES['''vocab_file'''] ) with open(self.vocab_file , '''w''' , encoding='''utf-8''' ) as vocab_writer: vocab_writer.write(''''''.join([x + '''\n''' for x in vocab_tokens] ) ) _UpperCamelCase = { '''do_resize''': True, '''size''': {'''height''': 18, '''width''': 18}, '''do_normalize''': True, '''image_mean''': [0.5, 0.5, 0.5], '''image_std''': [0.5, 0.5, 0.5], } _UpperCamelCase = os.path.join(self.tmpdirname , _A ) with open(self.image_processor_file , '''w''' , encoding='''utf-8''' ) as fp: json.dump(_A , _A ) def UpperCamelCase_ ( self : Tuple , **_A : Optional[Any] ): return BertTokenizer.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : List[Any] , **_A : Union[str, Any] ): return ViTImageProcessor.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : int ): shutil.rmtree(self.tmpdirname ) def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = [np.random.randint(255 , size=(3, 30, 400) , dtype=np.uinta )] _UpperCamelCase = [Image.fromarray(np.moveaxis(_A , 0 , -1 ) ) for x in image_inputs] return image_inputs def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = self.get_image_processor() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) processor.save_pretrained(self.tmpdirname ) _UpperCamelCase = VisionTextDualEncoderProcessor.from_pretrained(self.tmpdirname ) self.assertEqual(processor.tokenizer.get_vocab() , tokenizer.get_vocab() ) self.assertIsInstance(processor.tokenizer , (BertTokenizer, BertTokenizerFast) ) self.assertEqual(processor.image_processor.to_json_string() , image_processor.to_json_string() ) self.assertIsInstance(processor.image_processor , _A ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = VisionTextDualEncoderProcessor( tokenizer=self.get_tokenizer() , image_processor=self.get_image_processor() ) processor.save_pretrained(self.tmpdirname ) _UpperCamelCase = self.get_tokenizer(bos_token='''(BOS)''' , eos_token='''(EOS)''' ) _UpperCamelCase = self.get_image_processor(do_normalize=_A , padding_value=1.0 ) _UpperCamelCase = VisionTextDualEncoderProcessor.from_pretrained( self.tmpdirname , bos_token='''(BOS)''' , eos_token='''(EOS)''' , do_normalize=_A , padding_value=1.0 ) self.assertEqual(processor.tokenizer.get_vocab() , tokenizer_add_kwargs.get_vocab() ) self.assertIsInstance(processor.tokenizer , (BertTokenizer, BertTokenizerFast) ) self.assertEqual(processor.image_processor.to_json_string() , image_processor_add_kwargs.to_json_string() ) self.assertIsInstance(processor.image_processor , _A ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = image_processor(_A , return_tensors='''np''' ) _UpperCamelCase = processor(images=_A , return_tensors='''np''' ) for key in input_feat_extract.keys(): self.assertAlmostEqual(input_feat_extract[key].sum() , input_processor[key].sum() , delta=1e-2 ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = processor(text=_A ) _UpperCamelCase = tokenizer(_A ) for key in encoded_tok.keys(): self.assertListEqual(encoded_tok[key] , encoded_processor[key] ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = processor(text=_A , images=_A ) self.assertListEqual(list(inputs.keys() ) , ['''input_ids''', '''token_type_ids''', '''attention_mask''', '''pixel_values'''] ) # test if it raises when no input is passed with self.assertRaises(_A ): processor() def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = [[1, 4, 5, 8, 1, 0, 8], [3, 4, 3, 1, 1, 8, 9]] _UpperCamelCase = processor.batch_decode(_A ) _UpperCamelCase = tokenizer.batch_decode(_A ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = processor(text=_A , images=_A ) self.assertListEqual(list(inputs.keys() ) , processor.model_input_names )
71
1
import os from huggingface_hub.constants import HUGGINGFACE_HUB_CACHE, hf_cache_home _lowerCAmelCase = HUGGINGFACE_HUB_CACHE _lowerCAmelCase = "config.json" _lowerCAmelCase = "diffusion_pytorch_model.bin" _lowerCAmelCase = "diffusion_flax_model.msgpack" _lowerCAmelCase = "model.onnx" _lowerCAmelCase = "diffusion_pytorch_model.safetensors" _lowerCAmelCase = "weights.pb" _lowerCAmelCase = "https://huggingface.co" _lowerCAmelCase = default_cache_path _lowerCAmelCase = "diffusers_modules" _lowerCAmelCase = os.getenv("HF_MODULES_CACHE", os.path.join(hf_cache_home, "modules")) _lowerCAmelCase = ["fp16", "non-ema"] _lowerCAmelCase = ".self_attn"
71
def _snake_case ( __snake_case , __snake_case , __snake_case ): if n == 0: return 1 elif n % 2 == 1: return (binary_exponentiation(__snake_case , n - 1 , __snake_case ) * a) % mod else: _UpperCamelCase = binary_exponentiation(__snake_case , n / 2 , __snake_case ) return (b * b) % mod # a prime number _lowerCAmelCase = 701 _lowerCAmelCase = 1_000_000_000 _lowerCAmelCase = 10 # using binary exponentiation function, O(log(p)): print((a / b) % p == (a * binary_exponentiation(b, p - 2, p)) % p) print((a / b) % p == (a * b ** (p - 2)) % p)
71
1
# NOTE: This file is deprecated and will be removed in a future version. # It only exists so that temporarely `from diffusers.pipelines import DiffusionPipeline` works from ...utils import deprecate from ..controlnet.multicontrolnet import MultiControlNetModel # noqa: F401 from ..controlnet.pipeline_controlnet import StableDiffusionControlNetPipeline # noqa: F401 deprecate( "stable diffusion controlnet", "0.22.0", "Importing `StableDiffusionControlNetPipeline` or `MultiControlNetModel` from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_controlnet is deprecated. Please import `from diffusers import StableDiffusionControlNetPipeline` instead.", standard_warn=False, stacklevel=3, )
71
from math import cos, sin, sqrt, tau from audio_filters.iir_filter import IIRFilter def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 - _cos) / 2 _UpperCamelCase = 1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 + _cos) / 2 _UpperCamelCase = -1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = _sin / 2 _UpperCamelCase = 0 _UpperCamelCase = -ba _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 1 - alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([ba, ba, ba] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = 1 + alpha * big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha * big_a _UpperCamelCase = 1 + alpha / big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha / big_a _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (pmc + aaa) _UpperCamelCase = 2 * big_a * mpc _UpperCamelCase = big_a * (pmc - aaa) _UpperCamelCase = ppmc + aaa _UpperCamelCase = -2 * pmpc _UpperCamelCase = ppmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (ppmc + aaa) _UpperCamelCase = -2 * big_a * pmpc _UpperCamelCase = big_a * (ppmc - aaa) _UpperCamelCase = pmc + aaa _UpperCamelCase = 2 * mpc _UpperCamelCase = pmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt
71
1
from typing import TYPE_CHECKING from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_tokenizers_available, is_torch_available _lowerCAmelCase = { "configuration_nezha": ["NEZHA_PRETRAINED_CONFIG_ARCHIVE_MAP", "NezhaConfig"], } try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "NEZHA_PRETRAINED_MODEL_ARCHIVE_LIST", "NezhaForNextSentencePrediction", "NezhaForMaskedLM", "NezhaForPreTraining", "NezhaForMultipleChoice", "NezhaForQuestionAnswering", "NezhaForSequenceClassification", "NezhaForTokenClassification", "NezhaModel", "NezhaPreTrainedModel", ] if TYPE_CHECKING: from .configuration_nezha import NEZHA_PRETRAINED_CONFIG_ARCHIVE_MAP, NezhaConfig try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_nezha import ( NEZHA_PRETRAINED_MODEL_ARCHIVE_LIST, NezhaForMaskedLM, NezhaForMultipleChoice, NezhaForNextSentencePrediction, NezhaForPreTraining, NezhaForQuestionAnswering, NezhaForSequenceClassification, NezhaForTokenClassification, NezhaModel, NezhaPreTrainedModel, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
from ...configuration_utils import PretrainedConfig from ...utils import logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "EleutherAI/gpt-neox-20b": "https://huggingface.co/EleutherAI/gpt-neox-20b/resolve/main/config.json", # See all GPTNeoX models at https://huggingface.co/models?filter=gpt_neox } class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "gpt_neox" def __init__( self : Union[str, Any] , _A : Union[str, Any]=5_0432 , _A : List[Any]=6144 , _A : int=44 , _A : int=64 , _A : Optional[Any]=2_4576 , _A : Any="gelu" , _A : Tuple=0.25 , _A : Union[str, Any]=1_0000 , _A : Tuple=0.0 , _A : Any=0.0 , _A : int=0.1 , _A : List[str]=2048 , _A : Dict=0.02 , _A : Optional[Any]=1e-5 , _A : Tuple=True , _A : List[Any]=0 , _A : Optional[int]=2 , _A : Optional[int]=False , _A : List[Any]=True , _A : Any=None , **_A : Any , ): super().__init__(bos_token_id=_A , eos_token_id=_A , **_A ) _UpperCamelCase = vocab_size _UpperCamelCase = max_position_embeddings _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = rotary_pct _UpperCamelCase = rotary_emb_base _UpperCamelCase = attention_dropout _UpperCamelCase = hidden_dropout _UpperCamelCase = classifier_dropout _UpperCamelCase = initializer_range _UpperCamelCase = layer_norm_eps _UpperCamelCase = use_cache _UpperCamelCase = tie_word_embeddings _UpperCamelCase = use_parallel_residual _UpperCamelCase = rope_scaling self._rope_scaling_validation() if self.hidden_size % self.num_attention_heads != 0: raise ValueError( '''The hidden size is not divisble by the number of attention heads! Make sure to update them!''' ) def UpperCamelCase_ ( self : str ): if self.rope_scaling is None: return if not isinstance(self.rope_scaling , _A ) or len(self.rope_scaling ) != 2: raise ValueError( '''`rope_scaling` must be a dictionary with with two fields, `name` and `factor`, ''' F"""got {self.rope_scaling}""" ) _UpperCamelCase = self.rope_scaling.get('''type''' , _A ) _UpperCamelCase = self.rope_scaling.get('''factor''' , _A ) if rope_scaling_type is None or rope_scaling_type not in ["linear", "dynamic"]: raise ValueError( F"""`rope_scaling`'s name field must be one of ['linear', 'dynamic'], got {rope_scaling_type}""" ) if rope_scaling_factor is None or not isinstance(_A , _A ) or rope_scaling_factor <= 1.0: raise ValueError(F"""`rope_scaling`'s factor field must be an float > 1, got {rope_scaling_factor}""" )
71
1
import gc import random import unittest import numpy as np import torch from PIL import Image from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer from diffusers import AutoencoderKL, PNDMScheduler, StableDiffusionInpaintPipeline, UNetaDConditionModel from diffusers.utils import floats_tensor, load_image, load_numpy, torch_device from diffusers.utils.testing_utils import enable_full_determinism, require_torch_gpu, slow from ..pipeline_params import TEXT_GUIDED_IMAGE_INPAINTING_BATCH_PARAMS, TEXT_GUIDED_IMAGE_INPAINTING_PARAMS from ..test_pipelines_common import PipelineKarrasSchedulerTesterMixin, PipelineLatentTesterMixin, PipelineTesterMixin enable_full_determinism() class lowerCAmelCase_ ( __lowercase, __lowercase, __lowercase, unittest.TestCase ): UpperCAmelCase = StableDiffusionInpaintPipeline UpperCAmelCase = TEXT_GUIDED_IMAGE_INPAINTING_PARAMS UpperCAmelCase = TEXT_GUIDED_IMAGE_INPAINTING_BATCH_PARAMS UpperCAmelCase = frozenset( [] ) # TO-DO: update image_params once pipeline is refactored with VaeImageProcessor.preprocess UpperCAmelCase = frozenset([] ) def UpperCamelCase_ ( self : Optional[int] ): torch.manual_seed(0 ) _UpperCamelCase = UNetaDConditionModel( block_out_channels=(32, 64) , layers_per_block=2 , sample_size=32 , in_channels=9 , out_channels=4 , down_block_types=('''DownBlock2D''', '''CrossAttnDownBlock2D''') , up_block_types=('''CrossAttnUpBlock2D''', '''UpBlock2D''') , cross_attention_dim=32 , attention_head_dim=(2, 4) , use_linear_projection=_A , ) _UpperCamelCase = PNDMScheduler(skip_prk_steps=_A ) torch.manual_seed(0 ) _UpperCamelCase = AutoencoderKL( block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=['''DownEncoderBlock2D''', '''DownEncoderBlock2D'''] , up_block_types=['''UpDecoderBlock2D''', '''UpDecoderBlock2D'''] , latent_channels=4 , sample_size=128 , ) torch.manual_seed(0 ) _UpperCamelCase = CLIPTextConfig( bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-05 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1000 , hidden_act='''gelu''' , projection_dim=512 , ) _UpperCamelCase = CLIPTextModel(_A ) _UpperCamelCase = CLIPTokenizer.from_pretrained('''hf-internal-testing/tiny-random-clip''' ) _UpperCamelCase = { '''unet''': unet, '''scheduler''': scheduler, '''vae''': vae, '''text_encoder''': text_encoder, '''tokenizer''': tokenizer, '''safety_checker''': None, '''feature_extractor''': None, } return components def UpperCamelCase_ ( self : Union[str, Any] , _A : int , _A : Optional[int]=0 ): # TODO: use tensor inputs instead of PIL, this is here just to leave the old expected_slices untouched _UpperCamelCase = floats_tensor((1, 3, 32, 32) , rng=random.Random(_A ) ).to(_A ) _UpperCamelCase = image.cpu().permute(0 , 2 , 3 , 1 )[0] _UpperCamelCase = Image.fromarray(np.uinta(_A ) ).convert('''RGB''' ).resize((64, 64) ) _UpperCamelCase = Image.fromarray(np.uinta(image + 4 ) ).convert('''RGB''' ).resize((64, 64) ) if str(_A ).startswith('''mps''' ): _UpperCamelCase = torch.manual_seed(_A ) else: _UpperCamelCase = torch.Generator(device=_A ).manual_seed(_A ) _UpperCamelCase = { '''prompt''': '''A painting of a squirrel eating a burger''', '''image''': init_image, '''mask_image''': mask_image, '''generator''': generator, '''num_inference_steps''': 2, '''guidance_scale''': 6.0, '''output_type''': '''numpy''', } return inputs def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = '''cpu''' # ensure determinism for the device-dependent torch.Generator _UpperCamelCase = self.get_dummy_components() _UpperCamelCase = StableDiffusionInpaintPipeline(**_A ) _UpperCamelCase = sd_pipe.to(_A ) sd_pipe.set_progress_bar_config(disable=_A ) _UpperCamelCase = self.get_dummy_inputs(_A ) _UpperCamelCase = sd_pipe(**_A ).images _UpperCamelCase = image[0, -3:, -3:, -1] assert image.shape == (1, 64, 64, 3) _UpperCamelCase = np.array([0.4727, 0.5735, 0.3941, 0.5446, 0.5926, 0.4394, 0.5062, 0.4654, 0.4476] ) assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-2 def UpperCamelCase_ ( self : Optional[int] ): super().test_inference_batch_single_identical(expected_max_diff=3e-3 ) @slow @require_torch_gpu class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : List[str] ): # clean up the VRAM after each test super().tearDown() gc.collect() torch.cuda.empty_cache() def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = load_image( '''https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main''' '''/sd2-inpaint/init_image.png''' ) _UpperCamelCase = load_image( '''https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd2-inpaint/mask.png''' ) _UpperCamelCase = load_numpy( '''https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd2-inpaint''' '''/yellow_cat_sitting_on_a_park_bench.npy''' ) _UpperCamelCase = '''stabilityai/stable-diffusion-2-inpainting''' _UpperCamelCase = StableDiffusionInpaintPipeline.from_pretrained(_A , safety_checker=_A ) pipe.to(_A ) pipe.set_progress_bar_config(disable=_A ) pipe.enable_attention_slicing() _UpperCamelCase = '''Face of a yellow cat, high resolution, sitting on a park bench''' _UpperCamelCase = torch.manual_seed(0 ) _UpperCamelCase = pipe( prompt=_A , image=_A , mask_image=_A , generator=_A , output_type='''np''' , ) _UpperCamelCase = output.images[0] assert image.shape == (512, 512, 3) assert np.abs(expected_image - image ).max() < 9e-3 def UpperCamelCase_ ( self : int ): _UpperCamelCase = load_image( '''https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main''' '''/sd2-inpaint/init_image.png''' ) _UpperCamelCase = load_image( '''https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd2-inpaint/mask.png''' ) _UpperCamelCase = load_numpy( '''https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd2-inpaint''' '''/yellow_cat_sitting_on_a_park_bench_fp16.npy''' ) _UpperCamelCase = '''stabilityai/stable-diffusion-2-inpainting''' _UpperCamelCase = StableDiffusionInpaintPipeline.from_pretrained( _A , torch_dtype=torch.floataa , safety_checker=_A , ) pipe.to(_A ) pipe.set_progress_bar_config(disable=_A ) pipe.enable_attention_slicing() _UpperCamelCase = '''Face of a yellow cat, high resolution, sitting on a park bench''' _UpperCamelCase = torch.manual_seed(0 ) _UpperCamelCase = pipe( prompt=_A , image=_A , mask_image=_A , generator=_A , output_type='''np''' , ) _UpperCamelCase = output.images[0] assert image.shape == (512, 512, 3) assert np.abs(expected_image - image ).max() < 5e-1 def UpperCamelCase_ ( self : List[Any] ): torch.cuda.empty_cache() torch.cuda.reset_max_memory_allocated() torch.cuda.reset_peak_memory_stats() _UpperCamelCase = load_image( '''https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main''' '''/sd2-inpaint/init_image.png''' ) _UpperCamelCase = load_image( '''https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd2-inpaint/mask.png''' ) _UpperCamelCase = '''stabilityai/stable-diffusion-2-inpainting''' _UpperCamelCase = PNDMScheduler.from_pretrained(_A , subfolder='''scheduler''' ) _UpperCamelCase = StableDiffusionInpaintPipeline.from_pretrained( _A , safety_checker=_A , scheduler=_A , torch_dtype=torch.floataa , ) pipe.to(_A ) pipe.set_progress_bar_config(disable=_A ) pipe.enable_attention_slicing(1 ) pipe.enable_sequential_cpu_offload() _UpperCamelCase = '''Face of a yellow cat, high resolution, sitting on a park bench''' _UpperCamelCase = torch.manual_seed(0 ) _UpperCamelCase = pipe( prompt=_A , image=_A , mask_image=_A , generator=_A , num_inference_steps=2 , output_type='''np''' , ) _UpperCamelCase = torch.cuda.max_memory_allocated() # make sure that less than 2.65 GB is allocated assert mem_bytes < 2.65 * 10**9
71
from ..utils import DummyObject, requires_backends class lowerCAmelCase_ ( metaclass=__lowercase ): UpperCAmelCase = ["keras_nlp"] def __init__( self : Any , *_A : Dict , **_A : List[str] ): requires_backends(self , ['''keras_nlp'''] )
71
1
import logging import os import sys from dataclasses import dataclass, field from itertools import chain from typing import Optional, Union import datasets import numpy as np import torch from datasets import load_dataset import transformers from transformers import ( AutoConfig, AutoModelForMultipleChoice, AutoTokenizer, HfArgumentParser, Trainer, TrainingArguments, default_data_collator, set_seed, ) from transformers.tokenization_utils_base import PreTrainedTokenizerBase from transformers.trainer_utils import get_last_checkpoint from transformers.utils import PaddingStrategy, check_min_version, send_example_telemetry # Will error if the minimal version of Transformers is not installed. Remove at your own risks. check_min_version("4.31.0") _lowerCAmelCase = logging.getLogger(__name__) @dataclass class lowerCAmelCase_ : UpperCAmelCase = field( metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Pretrained config name or path if not the same as model_name"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"}, ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Whether to use one of the fast tokenizer (backed by the tokenizers library) or not."}, ) UpperCAmelCase = field( default="main", metadata={"help": "The specific model version to use (can be a branch name, tag name or commit id)."}, ) UpperCAmelCase = field( default=__lowercase, metadata={ "help": ( "Will use the token generated when running `huggingface-cli login` (necessary to use this script " "with private models)." ) }, ) @dataclass class lowerCAmelCase_ : UpperCAmelCase = field(default=__lowercase, metadata={"help": "The input training data file (a text file)."} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "An optional input evaluation data file to evaluate the perplexity on (a text file)."}, ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Overwrite the cached training and evaluation sets"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "The number of processes to use for the preprocessing."}, ) UpperCAmelCase = field( default=__lowercase, metadata={ "help": ( "The maximum total input sequence length after tokenization. If passed, sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field( default=__lowercase, metadata={ "help": ( "Whether to pad all samples to the maximum sentence length. " "If False, will pad the samples dynamically when batching to the maximum length in the batch. More " "efficient on GPU but very bad for TPU." ) }, ) UpperCAmelCase = field( default=__lowercase, metadata={ "help": ( "For debugging purposes or quicker training, truncate the number of training examples to this " "value if set." ) }, ) UpperCAmelCase = field( default=__lowercase, metadata={ "help": ( "For debugging purposes or quicker training, truncate the number of evaluation examples to this " "value if set." ) }, ) def UpperCamelCase_ ( self : Optional[int] ): if self.train_file is not None: _UpperCamelCase = self.train_file.split('''.''' )[-1] assert extension in ["csv", "json"], "`train_file` should be a csv or a json file." if self.validation_file is not None: _UpperCamelCase = self.validation_file.split('''.''' )[-1] assert extension in ["csv", "json"], "`validation_file` should be a csv or a json file." @dataclass class lowerCAmelCase_ : UpperCAmelCase = 42 UpperCAmelCase = True UpperCAmelCase = None UpperCAmelCase = None def __call__( self : Union[str, Any] , _A : Any ): _UpperCamelCase = '''label''' if '''label''' in features[0].keys() else '''labels''' _UpperCamelCase = [feature.pop(_A ) for feature in features] _UpperCamelCase = len(_A ) _UpperCamelCase = len(features[0]['''input_ids'''] ) _UpperCamelCase = [ [{k: v[i] for k, v in feature.items()} for i in range(_A )] for feature in features ] _UpperCamelCase = list(chain(*_A ) ) _UpperCamelCase = self.tokenizer.pad( _A , padding=self.padding , max_length=self.max_length , pad_to_multiple_of=self.pad_to_multiple_of , return_tensors='''pt''' , ) # Un-flatten _UpperCamelCase = {k: v.view(_A , _A , -1 ) for k, v in batch.items()} # Add back labels _UpperCamelCase = torch.tensor(_A , dtype=torch.intaa ) return batch def _snake_case ( ): # See all possible arguments in src/transformers/training_args.py # or by passing the --help flag to this script. # We now keep distinct sets of args, for a cleaner separation of concerns. _UpperCamelCase = HfArgumentParser((ModelArguments, DataTrainingArguments, TrainingArguments) ) if len(sys.argv ) == 2 and sys.argv[1].endswith('''.json''' ): # If we pass only one argument to the script and it's the path to a json file, # let's parse it to get our arguments. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1] ) ) else: _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = parser.parse_args_into_dataclasses() # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The # information sent is the one passed as arguments along with your Python/PyTorch versions. send_example_telemetry('''run_swag''' , __snake_case , __snake_case ) # Setup logging logging.basicConfig( format='''%(asctime)s - %(levelname)s - %(name)s - %(message)s''' , datefmt='''%m/%d/%Y %H:%M:%S''' , handlers=[logging.StreamHandler(sys.stdout )] , ) if training_args.should_log: # The default of training_args.log_level is passive, so we set log level at info here to have that default. transformers.utils.logging.set_verbosity_info() _UpperCamelCase = training_args.get_process_log_level() logger.setLevel(__snake_case ) datasets.utils.logging.set_verbosity(__snake_case ) transformers.utils.logging.set_verbosity(__snake_case ) transformers.utils.logging.enable_default_handler() transformers.utils.logging.enable_explicit_format() # Log on each process the small summary: logger.warning( f"""Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}""" + f"""distributed training: {bool(training_args.local_rank != -1 )}, 16-bits training: {training_args.fpaa}""" ) logger.info(f"""Training/evaluation parameters {training_args}""" ) # Detecting last checkpoint. _UpperCamelCase = None if os.path.isdir(training_args.output_dir ) and training_args.do_train and not training_args.overwrite_output_dir: _UpperCamelCase = get_last_checkpoint(training_args.output_dir ) if last_checkpoint is None and len(os.listdir(training_args.output_dir ) ) > 0: raise ValueError( f"""Output directory ({training_args.output_dir}) already exists and is not empty. """ '''Use --overwrite_output_dir to overcome.''' ) elif last_checkpoint is not None and training_args.resume_from_checkpoint is None: logger.info( f"""Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change """ '''the `--output_dir` or add `--overwrite_output_dir` to train from scratch.''' ) # Set seed before initializing model. set_seed(training_args.seed ) # Get the datasets: you can either provide your own CSV/JSON/TXT training and evaluation files (see below) # or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/ # (the dataset will be downloaded automatically from the datasets Hub). # For CSV/JSON files, this script will use the column called 'text' or the first column if no column called # 'text' is found. You can easily tweak this behavior (see below). # In distributed training, the load_dataset function guarantee that only one local process can concurrently # download the dataset. if data_args.train_file is not None or data_args.validation_file is not None: _UpperCamelCase = {} if data_args.train_file is not None: _UpperCamelCase = data_args.train_file if data_args.validation_file is not None: _UpperCamelCase = data_args.validation_file _UpperCamelCase = data_args.train_file.split('''.''' )[-1] _UpperCamelCase = load_dataset( __snake_case , data_files=__snake_case , cache_dir=model_args.cache_dir , use_auth_token=True if model_args.use_auth_token else None , ) else: # Downloading and loading the swag dataset from the hub. _UpperCamelCase = load_dataset( '''swag''' , '''regular''' , cache_dir=model_args.cache_dir , use_auth_token=True if model_args.use_auth_token else None , ) # See more about loading any type of standard or custom dataset (from files, python dict, pandas DataFrame, etc) at # https://huggingface.co/docs/datasets/loading_datasets.html. # Load pretrained model and tokenizer # Distributed training: # The .from_pretrained methods guarantee that only one local process can concurrently # download model & vocab. _UpperCamelCase = AutoConfig.from_pretrained( model_args.config_name if model_args.config_name else model_args.model_name_or_path , cache_dir=model_args.cache_dir , revision=model_args.model_revision , use_auth_token=True if model_args.use_auth_token else None , ) _UpperCamelCase = AutoTokenizer.from_pretrained( model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path , cache_dir=model_args.cache_dir , use_fast=model_args.use_fast_tokenizer , revision=model_args.model_revision , use_auth_token=True if model_args.use_auth_token else None , ) _UpperCamelCase = AutoModelForMultipleChoice.from_pretrained( model_args.model_name_or_path , from_tf=bool('''.ckpt''' in model_args.model_name_or_path ) , config=__snake_case , cache_dir=model_args.cache_dir , revision=model_args.model_revision , use_auth_token=True if model_args.use_auth_token else None , ) # When using your own dataset or a different dataset from swag, you will probably need to change this. _UpperCamelCase = [f"""ending{i}""" for i in range(4 )] _UpperCamelCase = '''sent1''' _UpperCamelCase = '''sent2''' if data_args.max_seq_length is None: _UpperCamelCase = tokenizer.model_max_length if max_seq_length > 1024: logger.warning( '''The chosen tokenizer supports a `model_max_length` that is longer than the default `block_size` value''' ''' of 1024. If you would like to use a longer `block_size` up to `tokenizer.model_max_length` you can''' ''' override this default with `--block_size xxx`.''' ) _UpperCamelCase = 1024 else: if data_args.max_seq_length > tokenizer.model_max_length: logger.warning( f"""The max_seq_length passed ({data_args.max_seq_length}) is larger than the maximum length for the""" f"""model ({tokenizer.model_max_length}). Using max_seq_length={tokenizer.model_max_length}.""" ) _UpperCamelCase = min(data_args.max_seq_length , tokenizer.model_max_length ) # Preprocessing the datasets. def preprocess_function(__snake_case ): _UpperCamelCase = [[context] * 4 for context in examples[context_name]] _UpperCamelCase = examples[question_header_name] _UpperCamelCase = [ [f"""{header} {examples[end][i]}""" for end in ending_names] for i, header in enumerate(__snake_case ) ] # Flatten out _UpperCamelCase = list(chain(*__snake_case ) ) _UpperCamelCase = list(chain(*__snake_case ) ) # Tokenize _UpperCamelCase = tokenizer( __snake_case , __snake_case , truncation=__snake_case , max_length=__snake_case , padding='''max_length''' if data_args.pad_to_max_length else False , ) # Un-flatten return {k: [v[i : i + 4] for i in range(0 , len(__snake_case ) , 4 )] for k, v in tokenized_examples.items()} if training_args.do_train: if "train" not in raw_datasets: raise ValueError('''--do_train requires a train dataset''' ) _UpperCamelCase = raw_datasets['''train'''] if data_args.max_train_samples is not None: _UpperCamelCase = min(len(__snake_case ) , data_args.max_train_samples ) _UpperCamelCase = train_dataset.select(range(__snake_case ) ) with training_args.main_process_first(desc='''train dataset map pre-processing''' ): _UpperCamelCase = train_dataset.map( __snake_case , batched=__snake_case , num_proc=data_args.preprocessing_num_workers , load_from_cache_file=not data_args.overwrite_cache , ) if training_args.do_eval: if "validation" not in raw_datasets: raise ValueError('''--do_eval requires a validation dataset''' ) _UpperCamelCase = raw_datasets['''validation'''] if data_args.max_eval_samples is not None: _UpperCamelCase = min(len(__snake_case ) , data_args.max_eval_samples ) _UpperCamelCase = eval_dataset.select(range(__snake_case ) ) with training_args.main_process_first(desc='''validation dataset map pre-processing''' ): _UpperCamelCase = eval_dataset.map( __snake_case , batched=__snake_case , num_proc=data_args.preprocessing_num_workers , load_from_cache_file=not data_args.overwrite_cache , ) # Data collator _UpperCamelCase = ( default_data_collator if data_args.pad_to_max_length else DataCollatorForMultipleChoice(tokenizer=__snake_case , pad_to_multiple_of=8 if training_args.fpaa else None ) ) # Metric def compute_metrics(__snake_case ): _UpperCamelCase , _UpperCamelCase = eval_predictions _UpperCamelCase = np.argmax(__snake_case , axis=1 ) return {"accuracy": (preds == label_ids).astype(np.floataa ).mean().item()} # Initialize our Trainer _UpperCamelCase = Trainer( model=__snake_case , args=__snake_case , train_dataset=train_dataset if training_args.do_train else None , eval_dataset=eval_dataset if training_args.do_eval else None , tokenizer=__snake_case , data_collator=__snake_case , compute_metrics=__snake_case , ) # Training if training_args.do_train: _UpperCamelCase = None if training_args.resume_from_checkpoint is not None: _UpperCamelCase = training_args.resume_from_checkpoint elif last_checkpoint is not None: _UpperCamelCase = last_checkpoint _UpperCamelCase = trainer.train(resume_from_checkpoint=__snake_case ) trainer.save_model() # Saves the tokenizer too for easy upload _UpperCamelCase = train_result.metrics _UpperCamelCase = ( data_args.max_train_samples if data_args.max_train_samples is not None else len(__snake_case ) ) _UpperCamelCase = min(__snake_case , len(__snake_case ) ) trainer.log_metrics('''train''' , __snake_case ) trainer.save_metrics('''train''' , __snake_case ) trainer.save_state() # Evaluation if training_args.do_eval: logger.info('''*** Evaluate ***''' ) _UpperCamelCase = trainer.evaluate() _UpperCamelCase = data_args.max_eval_samples if data_args.max_eval_samples is not None else len(__snake_case ) _UpperCamelCase = min(__snake_case , len(__snake_case ) ) trainer.log_metrics('''eval''' , __snake_case ) trainer.save_metrics('''eval''' , __snake_case ) _UpperCamelCase = { '''finetuned_from''': model_args.model_name_or_path, '''tasks''': '''multiple-choice''', '''dataset_tags''': '''swag''', '''dataset_args''': '''regular''', '''dataset''': '''SWAG''', '''language''': '''en''', } if training_args.push_to_hub: trainer.push_to_hub(**__snake_case ) else: trainer.create_model_card(**__snake_case ) def _snake_case ( __snake_case ): # For xla_spawn (TPUs) main() if __name__ == "__main__": main()
71
from typing import Optional, Tuple, Union import tensorflow as tf from ...activations_tf import ACTaFN from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward from ...modeling_tf_outputs import ( TFBaseModelOutputWithNoAttention, TFBaseModelOutputWithPoolingAndNoAttention, TFSequenceClassifierOutput, ) from ...modeling_tf_utils import TFPreTrainedModel, TFSequenceClassificationLoss, keras_serializable, unpack_inputs from ...tf_utils import shape_list from ...utils import logging from .configuration_regnet import RegNetConfig _lowerCAmelCase = logging.get_logger(__name__) # General docstring _lowerCAmelCase = "RegNetConfig" # Base docstring _lowerCAmelCase = "facebook/regnet-y-040" _lowerCAmelCase = [1, 1_088, 7, 7] # Image classification docstring _lowerCAmelCase = "facebook/regnet-y-040" _lowerCAmelCase = "tabby, tabby cat" _lowerCAmelCase = [ "facebook/regnet-y-040", # See all regnet models at https://huggingface.co/models?filter=regnet ] class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : str , _A : int , _A : int = 3 , _A : int = 1 , _A : int = 1 , _A : Optional[str] = "relu" , **_A : Any , ): super().__init__(**_A ) # The padding and conv has been verified in # https://colab.research.google.com/gist/sayakpaul/854bc10eeaf21c9ee2119e0b9f3841a7/scratchpad.ipynb _UpperCamelCase = tf.keras.layers.ZeroPaddingaD(padding=kernel_size // 2 ) _UpperCamelCase = tf.keras.layers.ConvaD( filters=_A , kernel_size=_A , strides=_A , padding='''VALID''' , groups=_A , use_bias=_A , name='''convolution''' , ) _UpperCamelCase = tf.keras.layers.BatchNormalization(epsilon=1e-5 , momentum=0.9 , name='''normalization''' ) _UpperCamelCase = ACTaFN[activation] if activation is not None else tf.identity def UpperCamelCase_ ( self : Any , _A : Any ): _UpperCamelCase = self.convolution(self.padding(_A ) ) _UpperCamelCase = self.normalization(_A ) _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Optional[Any] , _A : RegNetConfig , **_A : Any ): super().__init__(**_A ) _UpperCamelCase = config.num_channels _UpperCamelCase = TFRegNetConvLayer( out_channels=config.embedding_size , kernel_size=3 , stride=2 , activation=config.hidden_act , name='''embedder''' , ) def UpperCamelCase_ ( self : List[str] , _A : Optional[int] ): _UpperCamelCase = shape_list(_A )[1] if tf.executing_eagerly() and num_channels != self.num_channels: raise ValueError( '''Make sure that the channel dimension of the pixel values match with the one set in the configuration.''' ) # When running on CPU, `tf.keras.layers.Conv2D` doesn't support `NCHW` format. # So change the input format from `NCHW` to `NHWC`. # shape = (batch_size, in_height, in_width, in_channels=num_channels) _UpperCamelCase = tf.transpose(_A , perm=(0, 2, 3, 1) ) _UpperCamelCase = self.embedder(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : str , _A : int , _A : int = 2 , **_A : Optional[Any] ): super().__init__(**_A ) _UpperCamelCase = tf.keras.layers.ConvaD( filters=_A , kernel_size=1 , strides=_A , use_bias=_A , name='''convolution''' ) _UpperCamelCase = tf.keras.layers.BatchNormalization(epsilon=1e-5 , momentum=0.9 , name='''normalization''' ) def UpperCamelCase_ ( self : str , _A : tf.Tensor , _A : bool = False ): return self.normalization(self.convolution(_A ) , training=_A ) class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Dict , _A : int , _A : int , **_A : Dict ): super().__init__(**_A ) _UpperCamelCase = tf.keras.layers.GlobalAveragePoolingaD(keepdims=_A , name='''pooler''' ) _UpperCamelCase = [ tf.keras.layers.ConvaD(filters=_A , kernel_size=1 , activation='''relu''' , name='''attention.0''' ), tf.keras.layers.ConvaD(filters=_A , kernel_size=1 , activation='''sigmoid''' , name='''attention.2''' ), ] def UpperCamelCase_ ( self : List[str] , _A : List[Any] ): # [batch_size, h, w, num_channels] -> [batch_size, 1, 1, num_channels] _UpperCamelCase = self.pooler(_A ) for layer_module in self.attention: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = hidden_state * pooled return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , _A : int , _A : int , _A : int = 1 , **_A : str ): super().__init__(**_A ) _UpperCamelCase = in_channels != out_channels or stride != 1 _UpperCamelCase = max(1 , out_channels // config.groups_width ) _UpperCamelCase = ( TFRegNetShortCut(_A , stride=_A , name='''shortcut''' ) if should_apply_shortcut else tf.keras.layers.Activation('''linear''' , name='''shortcut''' ) ) # `self.layers` instead of `self.layer` because that is a reserved argument. _UpperCamelCase = [ TFRegNetConvLayer(_A , kernel_size=1 , activation=config.hidden_act , name='''layer.0''' ), TFRegNetConvLayer( _A , stride=_A , groups=_A , activation=config.hidden_act , name='''layer.1''' ), TFRegNetConvLayer(_A , kernel_size=1 , activation=_A , name='''layer.2''' ), ] _UpperCamelCase = ACTaFN[config.hidden_act] def UpperCamelCase_ ( self : Dict , _A : Tuple ): _UpperCamelCase = hidden_state for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = self.shortcut(_A ) hidden_state += residual _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , _A : int , _A : int , _A : int = 1 , **_A : int ): super().__init__(**_A ) _UpperCamelCase = in_channels != out_channels or stride != 1 _UpperCamelCase = max(1 , out_channels // config.groups_width ) _UpperCamelCase = ( TFRegNetShortCut(_A , stride=_A , name='''shortcut''' ) if should_apply_shortcut else tf.keras.layers.Activation('''linear''' , name='''shortcut''' ) ) _UpperCamelCase = [ TFRegNetConvLayer(_A , kernel_size=1 , activation=config.hidden_act , name='''layer.0''' ), TFRegNetConvLayer( _A , stride=_A , groups=_A , activation=config.hidden_act , name='''layer.1''' ), TFRegNetSELayer(_A , reduced_channels=int(round(in_channels / 4 ) ) , name='''layer.2''' ), TFRegNetConvLayer(_A , kernel_size=1 , activation=_A , name='''layer.3''' ), ] _UpperCamelCase = ACTaFN[config.hidden_act] def UpperCamelCase_ ( self : Tuple , _A : List[Any] ): _UpperCamelCase = hidden_state for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = self.shortcut(_A ) hidden_state += residual _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Tuple , _A : RegNetConfig , _A : int , _A : int , _A : int = 2 , _A : int = 2 , **_A : Union[str, Any] ): super().__init__(**_A ) _UpperCamelCase = TFRegNetXLayer if config.layer_type == '''x''' else TFRegNetYLayer _UpperCamelCase = [ # downsampling is done in the first layer with stride of 2 layer(_A , _A , _A , stride=_A , name='''layers.0''' ), *[layer(_A , _A , _A , name=F"""layers.{i+1}""" ) for i in range(depth - 1 )], ] def UpperCamelCase_ ( self : Union[str, Any] , _A : Optional[int] ): for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , **_A : List[str] ): super().__init__(**_A ) _UpperCamelCase = [] # based on `downsample_in_first_stage`, the first layer of the first stage may or may not downsample the input self.stages.append( TFRegNetStage( _A , config.embedding_size , config.hidden_sizes[0] , stride=2 if config.downsample_in_first_stage else 1 , depth=config.depths[0] , name='''stages.0''' , ) ) _UpperCamelCase = zip(config.hidden_sizes , config.hidden_sizes[1:] ) for i, ((in_channels, out_channels), depth) in enumerate(zip(_A , config.depths[1:] ) ): self.stages.append(TFRegNetStage(_A , _A , _A , depth=_A , name=F"""stages.{i+1}""" ) ) def UpperCamelCase_ ( self : Optional[int] , _A : tf.Tensor , _A : bool = False , _A : bool = True ): _UpperCamelCase = () if output_hidden_states else None for stage_module in self.stages: if output_hidden_states: _UpperCamelCase = hidden_states + (hidden_state,) _UpperCamelCase = stage_module(_A ) if output_hidden_states: _UpperCamelCase = hidden_states + (hidden_state,) if not return_dict: return tuple(v for v in [hidden_state, hidden_states] if v is not None ) return TFBaseModelOutputWithNoAttention(last_hidden_state=_A , hidden_states=_A ) @keras_serializable class lowerCAmelCase_ ( tf.keras.layers.Layer ): UpperCAmelCase = RegNetConfig def __init__( self : int , _A : Tuple , **_A : int ): super().__init__(**_A ) _UpperCamelCase = config _UpperCamelCase = TFRegNetEmbeddings(_A , name='''embedder''' ) _UpperCamelCase = TFRegNetEncoder(_A , name='''encoder''' ) _UpperCamelCase = tf.keras.layers.GlobalAveragePoolingaD(keepdims=_A , name='''pooler''' ) @unpack_inputs def UpperCamelCase_ ( self : Optional[int] , _A : tf.Tensor , _A : Optional[bool] = None , _A : Optional[bool] = None , _A : bool = False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.embedder(_A , training=_A ) _UpperCamelCase = self.encoder( _A , output_hidden_states=_A , return_dict=_A , training=_A ) _UpperCamelCase = encoder_outputs[0] _UpperCamelCase = self.pooler(_A ) # Change to NCHW output format have uniformity in the modules _UpperCamelCase = tf.transpose(_A , perm=(0, 3, 1, 2) ) _UpperCamelCase = tf.transpose(_A , perm=(0, 3, 1, 2) ) # Change the other hidden state outputs to NCHW as well if output_hidden_states: _UpperCamelCase = tuple([tf.transpose(_A , perm=(0, 3, 1, 2) ) for h in encoder_outputs[1]] ) if not return_dict: return (last_hidden_state, pooled_output) + encoder_outputs[1:] return TFBaseModelOutputWithPoolingAndNoAttention( last_hidden_state=_A , pooler_output=_A , hidden_states=hidden_states if output_hidden_states else encoder_outputs.hidden_states , ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = RegNetConfig UpperCAmelCase = "regnet" UpperCAmelCase = "pixel_values" @property def UpperCamelCase_ ( self : Tuple ): return {"pixel_values": tf.TensorSpec(shape=(None, self.config.num_channels, 224, 224) , dtype=tf.floataa )} _lowerCAmelCase = r"\n Parameters:\n This model is a Tensorflow\n [tf.keras.layers.Layer](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Layer) sub-class. Use it as a\n regular Tensorflow Module and refer to the Tensorflow documentation for all matter related to general usage and\n behavior.\n config ([`RegNetConfig`]): Model configuration class with all the parameters of the model.\n Initializing with a config file does not load the weights associated with the model, only the\n configuration. Check out the [`~TFPreTrainedModel.from_pretrained`] method to load the model weights.\n" _lowerCAmelCase = r"\n Args:\n pixel_values (`tf.Tensor` of shape `(batch_size, num_channels, height, width)`):\n Pixel values. Pixel values can be obtained using [`AutoImageProcessor`]. See\n [`ConveNextImageProcessor.__call__`] for details.\n output_hidden_states (`bool`, *optional*):\n Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for\n more detail.\n return_dict (`bool`, *optional*):\n Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.\n" @add_start_docstrings( "The bare RegNet model outputting raw features without any specific head on top.", __lowercase, ) class lowerCAmelCase_ ( __lowercase ): def __init__( self : List[Any] , _A : RegNetConfig , *_A : Optional[int] , **_A : Tuple ): super().__init__(_A , *_A , **_A ) _UpperCamelCase = TFRegNetMainLayer(_A , name='''regnet''' ) @unpack_inputs @add_start_docstrings_to_model_forward(_A ) @add_code_sample_docstrings( checkpoint=_CHECKPOINT_FOR_DOC , output_type=_A , config_class=_CONFIG_FOR_DOC , modality='''vision''' , expected_output=_EXPECTED_OUTPUT_SHAPE , ) def UpperCamelCase_ ( self : Any , _A : tf.Tensor , _A : Optional[bool] = None , _A : Optional[bool] = None , _A : Optional[int]=False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.regnet( pixel_values=_A , output_hidden_states=_A , return_dict=_A , training=_A , ) if not return_dict: return (outputs[0],) + outputs[1:] return TFBaseModelOutputWithPoolingAndNoAttention( last_hidden_state=outputs.last_hidden_state , pooler_output=outputs.pooler_output , hidden_states=outputs.hidden_states , ) @add_start_docstrings( "\n RegNet Model with an image classification head on top (a linear layer on top of the pooled features), e.g. for\n ImageNet.\n ", __lowercase, ) class lowerCAmelCase_ ( __lowercase, __lowercase ): def __init__( self : List[Any] , _A : RegNetConfig , *_A : Any , **_A : int ): super().__init__(_A , *_A , **_A ) _UpperCamelCase = config.num_labels _UpperCamelCase = TFRegNetMainLayer(_A , name='''regnet''' ) # classification head _UpperCamelCase = [ tf.keras.layers.Flatten(), tf.keras.layers.Dense(config.num_labels , name='''classifier.1''' ) if config.num_labels > 0 else tf.identity, ] @unpack_inputs @add_start_docstrings_to_model_forward(_A ) @add_code_sample_docstrings( checkpoint=_IMAGE_CLASS_CHECKPOINT , output_type=_A , config_class=_CONFIG_FOR_DOC , expected_output=_IMAGE_CLASS_EXPECTED_OUTPUT , ) def UpperCamelCase_ ( self : str , _A : tf.Tensor = None , _A : tf.Tensor = None , _A : bool = None , _A : bool = None , _A : Any=False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.regnet( _A , output_hidden_states=_A , return_dict=_A , training=_A ) _UpperCamelCase = outputs.pooler_output if return_dict else outputs[1] _UpperCamelCase = self.classifier[0](_A ) _UpperCamelCase = self.classifier[1](_A ) _UpperCamelCase = None if labels is None else self.hf_compute_loss(labels=_A , logits=_A ) if not return_dict: _UpperCamelCase = (logits,) + outputs[2:] return ((loss,) + output) if loss is not None else output return TFSequenceClassifierOutput(loss=_A , logits=_A , hidden_states=outputs.hidden_states )
71
1
import math def _snake_case ( __snake_case = 100 ): _UpperCamelCase = sum(i * i for i in range(1 , n + 1 ) ) _UpperCamelCase = int(math.pow(sum(range(1 , n + 1 ) ) , 2 ) ) return square_of_sum - sum_of_squares if __name__ == "__main__": print(f'{solution() = }')
71
from sklearn.metrics import mean_squared_error import datasets _lowerCAmelCase = "\\n@article{scikit-learn,\n title={Scikit-learn: Machine Learning in {P}ython},\n author={Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V.\n and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P.\n and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and\n Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E.},\n journal={Journal of Machine Learning Research},\n volume={12},\n pages={2825--2830},\n year={2011}\n}\n" _lowerCAmelCase = "\\nMean Squared Error(MSE) is the average of the square of difference between the predicted\nand actual values.\n" _lowerCAmelCase = "\nArgs:\n predictions: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Estimated target values.\n references: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Ground truth (correct) target values.\n sample_weight: array-like of shape (n_samples,), default=None\n Sample weights.\n multioutput: {\"raw_values\", \"uniform_average\"} or array-like of shape (n_outputs,), default=\"uniform_average\"\n Defines aggregating of multiple output values. Array-like value defines weights used to average errors.\n\n \"raw_values\" : Returns a full set of errors in case of multioutput input.\n\n \"uniform_average\" : Errors of all outputs are averaged with uniform weight.\n\n squared : bool, default=True\n If True returns MSE value, if False returns RMSE (Root Mean Squared Error) value.\n\nReturns:\n mse : mean squared error.\nExamples:\n\n >>> mse_metric = datasets.load_metric(\"mse\")\n >>> predictions = [2.5, 0.0, 2, 8]\n >>> references = [3, -0.5, 2, 7]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.375}\n >>> rmse_result = mse_metric.compute(predictions=predictions, references=references, squared=False)\n >>> print(rmse_result)\n {'mse': 0.6123724356957945}\n\n If you're using multi-dimensional lists, then set the config as follows :\n\n >>> mse_metric = datasets.load_metric(\"mse\", \"multilist\")\n >>> predictions = [[0.5, 1], [-1, 1], [7, -6]]\n >>> references = [[0, 2], [-1, 2], [8, -5]]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.7083333333333334}\n >>> results = mse_metric.compute(predictions=predictions, references=references, multioutput='raw_values')\n >>> print(results) # doctest: +NORMALIZE_WHITESPACE\n {'mse': array([0.41666667, 1. ])}\n" @datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION ) class lowerCAmelCase_ ( datasets.Metric ): def UpperCamelCase_ ( self : Optional[int] ): return datasets.MetricInfo( description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features(self._get_feature_types() ) , reference_urls=[ '''https://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html''' ] , ) def UpperCamelCase_ ( self : Dict ): if self.config_name == "multilist": return { "predictions": datasets.Sequence(datasets.Value('''float''' ) ), "references": datasets.Sequence(datasets.Value('''float''' ) ), } else: return { "predictions": datasets.Value('''float''' ), "references": datasets.Value('''float''' ), } def UpperCamelCase_ ( self : Any , _A : List[Any] , _A : List[str] , _A : Dict=None , _A : List[str]="uniform_average" , _A : int=True ): _UpperCamelCase = mean_squared_error( _A , _A , sample_weight=_A , multioutput=_A , squared=_A ) return {"mse": mse}
71
1
import inspect import unittest from transformers import RegNetConfig, is_flax_available from transformers.testing_utils import require_flax, slow from transformers.utils import cached_property, is_vision_available from ...test_configuration_common import ConfigTester from ...test_modeling_flax_common import FlaxModelTesterMixin, floats_tensor if is_flax_available(): import jax import jax.numpy as jnp from transformers.models.regnet.modeling_flax_regnet import FlaxRegNetForImageClassification, FlaxRegNetModel if is_vision_available(): from PIL import Image from transformers import AutoImageProcessor class lowerCAmelCase_ ( unittest.TestCase ): def __init__( self : Any , _A : List[Any] , _A : List[str]=3 , _A : Union[str, Any]=32 , _A : List[str]=3 , _A : Any=10 , _A : str=[10, 20, 30, 40] , _A : Any=[1, 1, 2, 1] , _A : Any=True , _A : str=True , _A : Dict="relu" , _A : Tuple=3 , _A : List[Any]=None , ): _UpperCamelCase = parent _UpperCamelCase = batch_size _UpperCamelCase = image_size _UpperCamelCase = num_channels _UpperCamelCase = embeddings_size _UpperCamelCase = hidden_sizes _UpperCamelCase = depths _UpperCamelCase = is_training _UpperCamelCase = use_labels _UpperCamelCase = hidden_act _UpperCamelCase = num_labels _UpperCamelCase = scope _UpperCamelCase = len(_A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] ) _UpperCamelCase = self.get_config() return config, pixel_values def UpperCamelCase_ ( self : Union[str, Any] ): return RegNetConfig( num_channels=self.num_channels , embeddings_size=self.embeddings_size , hidden_sizes=self.hidden_sizes , depths=self.depths , hidden_act=self.hidden_act , num_labels=self.num_labels , image_size=self.image_size , ) def UpperCamelCase_ ( self : str , _A : Union[str, Any] , _A : str ): _UpperCamelCase = FlaxRegNetModel(config=_A ) _UpperCamelCase = model(_A ) # Output shape (b, c, h, w) self.parent.assertEqual( result.last_hidden_state.shape , (self.batch_size, self.hidden_sizes[-1], self.image_size // 32, self.image_size // 32) , ) def UpperCamelCase_ ( self : Optional[int] , _A : Union[str, Any] , _A : str ): _UpperCamelCase = self.num_labels _UpperCamelCase = FlaxRegNetForImageClassification(config=_A ) _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.prepare_config_and_inputs() _UpperCamelCase , _UpperCamelCase = config_and_inputs _UpperCamelCase = {'''pixel_values''': pixel_values} return config, inputs_dict @require_flax class lowerCAmelCase_ ( __lowercase, unittest.TestCase ): UpperCAmelCase = (FlaxRegNetModel, FlaxRegNetForImageClassification) if is_flax_available() else () UpperCAmelCase = False UpperCAmelCase = False UpperCAmelCase = False def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = FlaxRegNetModelTester(self ) _UpperCamelCase = ConfigTester(self , config_class=_A , has_text_modality=_A ) def UpperCamelCase_ ( self : Any ): self.create_and_test_config_common_properties() self.config_tester.create_and_test_config_to_json_string() self.config_tester.create_and_test_config_to_json_file() self.config_tester.create_and_test_config_from_and_save_pretrained() self.config_tester.create_and_test_config_with_num_labels() self.config_tester.check_config_can_be_init_without_params() self.config_tester.check_config_arguments_init() def UpperCamelCase_ ( self : List[str] ): return def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model(*_A ) def UpperCamelCase_ ( self : str ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_image_classification(*_A ) @unittest.skip(reason='''RegNet does not use inputs_embeds''' ) def UpperCamelCase_ ( self : Union[str, Any] ): pass @unittest.skip(reason='''RegNet does not support input and output embeddings''' ) def UpperCamelCase_ ( self : Dict ): pass def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase , _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() for model_class in self.all_model_classes: _UpperCamelCase = model_class(_A ) _UpperCamelCase = inspect.signature(model.__call__ ) # signature.parameters is an OrderedDict => so arg_names order is deterministic _UpperCamelCase = [*signature.parameters.keys()] _UpperCamelCase = ['''pixel_values'''] self.assertListEqual(arg_names[:1] , _A ) def UpperCamelCase_ ( self : Optional[Any] ): def check_hidden_states_output(_A : Optional[int] , _A : Union[str, Any] , _A : Optional[int] ): _UpperCamelCase = model_class(_A ) _UpperCamelCase = model(**self._prepare_for_class(_A , _A ) ) _UpperCamelCase = outputs.encoder_hidden_states if config.is_encoder_decoder else outputs.hidden_states _UpperCamelCase = self.model_tester.num_stages self.assertEqual(len(_A ) , expected_num_stages + 1 ) _UpperCamelCase , _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() for model_class in self.all_model_classes: _UpperCamelCase = True check_hidden_states_output(_A , _A , _A ) # check that output_hidden_states also work using config del inputs_dict["output_hidden_states"] _UpperCamelCase = True check_hidden_states_output(_A , _A , _A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase , _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() for model_class in self.all_model_classes: with self.subTest(model_class.__name__ ): _UpperCamelCase = self._prepare_for_class(_A , _A ) _UpperCamelCase = model_class(_A ) @jax.jit def model_jitted(_A : Any , **_A : Dict ): return model(pixel_values=_A , **_A ) with self.subTest('''JIT Enabled''' ): _UpperCamelCase = model_jitted(**_A ).to_tuple() with self.subTest('''JIT Disabled''' ): with jax.disable_jit(): _UpperCamelCase = model_jitted(**_A ).to_tuple() self.assertEqual(len(_A ) , len(_A ) ) for jitted_output, output in zip(_A , _A ): self.assertEqual(jitted_output.shape , output.shape ) def _snake_case ( ): _UpperCamelCase = Image.open('''./tests/fixtures/tests_samples/COCO/000000039769.png''' ) return image @require_flax class lowerCAmelCase_ ( unittest.TestCase ): @cached_property def UpperCamelCase_ ( self : str ): return AutoImageProcessor.from_pretrained('''facebook/regnet-y-040''' ) if is_vision_available() else None @slow def UpperCamelCase_ ( self : str ): _UpperCamelCase = FlaxRegNetForImageClassification.from_pretrained('''facebook/regnet-y-040''' ) _UpperCamelCase = self.default_image_processor _UpperCamelCase = prepare_img() _UpperCamelCase = image_processor(images=_A , return_tensors='''np''' ) _UpperCamelCase = model(**_A ) # verify the logits _UpperCamelCase = (1, 1000) self.assertEqual(outputs.logits.shape , _A ) _UpperCamelCase = jnp.array([-0.4180, -1.5051, -3.4836] ) self.assertTrue(jnp.allclose(outputs.logits[0, :3] , _A , atol=1e-4 ) )
71
import os import re import shutil import sys import tempfile import unittest import black _lowerCAmelCase = os.path.abspath(os.path.dirname(os.path.dirname(os.path.dirname(__file__)))) sys.path.append(os.path.join(git_repo_path, "utils")) import check_copies # noqa: E402 # This is the reference code that will be used in the tests. # If DDPMSchedulerOutput is changed in scheduling_ddpm.py, this code needs to be manually updated. _lowerCAmelCase = " \"\"\"\n Output class for the scheduler's step function output.\n\n Args:\n prev_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n Computed sample (x_{t-1}) of previous timestep. `prev_sample` should be used as next model input in the\n denoising loop.\n pred_original_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n The predicted denoised sample (x_{0}) based on the model output from the current timestep.\n `pred_original_sample` can be used to preview progress or for guidance.\n \"\"\"\n\n prev_sample: torch.FloatTensor\n pred_original_sample: Optional[torch.FloatTensor] = None\n" class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = tempfile.mkdtemp() os.makedirs(os.path.join(self.diffusers_dir , '''schedulers/''' ) ) _UpperCamelCase = self.diffusers_dir shutil.copy( os.path.join(_A , '''src/diffusers/schedulers/scheduling_ddpm.py''' ) , os.path.join(self.diffusers_dir , '''schedulers/scheduling_ddpm.py''' ) , ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = '''src/diffusers''' shutil.rmtree(self.diffusers_dir ) def UpperCamelCase_ ( self : Union[str, Any] , _A : Tuple , _A : Optional[Any] , _A : Dict , _A : List[str]=None ): _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + class_code if overwrite_result is not None: _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + overwrite_result _UpperCamelCase = black.Mode(target_versions={black.TargetVersion.PYaa} , line_length=119 ) _UpperCamelCase = black.format_str(_A , mode=_A ) _UpperCamelCase = os.path.join(self.diffusers_dir , '''new_code.py''' ) with open(_A , '''w''' , newline='''\n''' ) as f: f.write(_A ) if overwrite_result is None: self.assertTrue(len(check_copies.is_copy_consistent(_A ) ) == 0 ) else: check_copies.is_copy_consistent(f.name , overwrite=_A ) with open(_A , '''r''' ) as f: self.assertTrue(f.read() , _A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = check_copies.find_code_in_diffusers('''schedulers.scheduling_ddpm.DDPMSchedulerOutput''' ) self.assertEqual(_A , _A ) def UpperCamelCase_ ( self : Optional[Any] ): # Base copy consistency self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , REFERENCE_CODE + '''\n''' , ) # With no empty line at the end self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , _A , ) # Copy consistency with rename self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , re.sub('''DDPM''' , '''Test''' , _A ) , ) # Copy consistency with a really long name _UpperCamelCase = '''TestClassWithAReallyLongNameBecauseSomePeopleLikeThatForSomeReason''' self.check_copy_consistency( F"""# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->{long_class_name}""" , F"""{long_class_name}SchedulerOutput""" , re.sub('''Bert''' , _A , _A ) , ) # Copy consistency with overwrite self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , _A , overwrite_result=re.sub('''DDPM''' , '''Test''' , _A ) , )
71
1
import numpy as np from cva import COLOR_BGR2GRAY, cvtColor, imread from numpy import array, uinta from PIL import Image from digital_image_processing import change_contrast as cc from digital_image_processing import convert_to_negative as cn from digital_image_processing import sepia as sp from digital_image_processing.dithering import burkes as bs from digital_image_processing.edge_detection import canny from digital_image_processing.filters import convolve as conv from digital_image_processing.filters import gaussian_filter as gg from digital_image_processing.filters import local_binary_pattern as lbp from digital_image_processing.filters import median_filter as med from digital_image_processing.filters import sobel_filter as sob from digital_image_processing.resize import resize as rs _lowerCAmelCase = imread(r"digital_image_processing/image_data/lena_small.jpg") _lowerCAmelCase = cvtColor(img, COLOR_BGR2GRAY) def _snake_case ( ): _UpperCamelCase = cn.convert_to_negative(__snake_case ) # assert negative_img array for at least one True assert negative_img.any() def _snake_case ( ): with Image.open('''digital_image_processing/image_data/lena_small.jpg''' ) as img: # Work around assertion for response assert str(cc.change_contrast(__snake_case , 110 ) ).startswith( '''<PIL.Image.Image image mode=RGB size=100x100 at''' ) def _snake_case ( ): _UpperCamelCase = canny.gen_gaussian_kernel(9 , sigma=1.4 ) # Assert ambiguous array assert resp.all() def _snake_case ( ): _UpperCamelCase = imread('''digital_image_processing/image_data/lena_small.jpg''' , 0 ) # assert ambiguous array for all == True assert canny_img.all() _UpperCamelCase = canny.canny(__snake_case ) # assert canny array for at least one True assert canny_array.any() def _snake_case ( ): assert gg.gaussian_filter(__snake_case , 5 , sigma=0.9 ).all() def _snake_case ( ): # laplace diagonals _UpperCamelCase = array([[0.25, 0.5, 0.25], [0.5, -3, 0.5], [0.25, 0.5, 0.25]] ) _UpperCamelCase = conv.img_convolve(__snake_case , __snake_case ).astype(__snake_case ) assert res.any() def _snake_case ( ): assert med.median_filter(__snake_case , 3 ).any() def _snake_case ( ): _UpperCamelCase , _UpperCamelCase = sob.sobel_filter(__snake_case ) assert grad.any() and theta.any() def _snake_case ( ): _UpperCamelCase = sp.make_sepia(__snake_case , 20 ) assert sepia.all() def _snake_case ( __snake_case = "digital_image_processing/image_data/lena_small.jpg" ): _UpperCamelCase = bs.Burkes(imread(__snake_case , 1 ) , 120 ) burkes.process() assert burkes.output_img.any() def _snake_case ( __snake_case = "digital_image_processing/image_data/lena_small.jpg" , ): _UpperCamelCase = rs.NearestNeighbour(imread(__snake_case , 1 ) , 400 , 200 ) nn.process() assert nn.output.any() def _snake_case ( ): _UpperCamelCase = '''digital_image_processing/image_data/lena.jpg''' # Reading the image and converting it to grayscale. _UpperCamelCase = imread(__snake_case , 0 ) # Test for get_neighbors_pixel function() return not None _UpperCamelCase = 0 _UpperCamelCase = 0 _UpperCamelCase = image[x_coordinate][y_coordinate] _UpperCamelCase = lbp.get_neighbors_pixel( __snake_case , __snake_case , __snake_case , __snake_case ) assert neighbors_pixels is not None # Test for local_binary_pattern function() # Create a numpy array as the same height and width of read image _UpperCamelCase = np.zeros((image.shape[0], image.shape[1]) ) # Iterating through the image and calculating the local binary pattern value # for each pixel. for i in range(0 , image.shape[0] ): for j in range(0 , image.shape[1] ): _UpperCamelCase = lbp.local_binary_value(__snake_case , __snake_case , __snake_case ) assert lbp_image.any()
71
from __future__ import annotations import math class lowerCAmelCase_ : def __init__( self : int , _A : int ): _UpperCamelCase = size # approximate the overall size of segment tree with given value _UpperCamelCase = [0 for i in range(0 , 4 * size )] # create array to store lazy update _UpperCamelCase = [0 for i in range(0 , 4 * size )] _UpperCamelCase = [0 for i in range(0 , 4 * size )] # flag for lazy update def UpperCamelCase_ ( self : str , _A : int ): return idx * 2 def UpperCamelCase_ ( self : Any , _A : int ): return idx * 2 + 1 def UpperCamelCase_ ( self : Union[str, Any] , _A : int , _A : int , _A : int , _A : list[int] ): if left_element == right_element: _UpperCamelCase = a[left_element - 1] else: _UpperCamelCase = (left_element + right_element) // 2 self.build(self.left(_A ) , _A , _A , _A ) self.build(self.right(_A ) , mid + 1 , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) def UpperCamelCase_ ( self : Tuple , _A : int , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return True if left_element >= a and right_element <= b: _UpperCamelCase = val if left_element != right_element: _UpperCamelCase = val _UpperCamelCase = val _UpperCamelCase = True _UpperCamelCase = True return True _UpperCamelCase = (left_element + right_element) // 2 self.update(self.left(_A ) , _A , _A , _A , _A , _A ) self.update(self.right(_A ) , mid + 1 , _A , _A , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) return True def UpperCamelCase_ ( self : Any , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return -math.inf if left_element >= a and right_element <= b: return self.segment_tree[idx] _UpperCamelCase = (left_element + right_element) // 2 _UpperCamelCase = self.query(self.left(_A ) , _A , _A , _A , _A ) _UpperCamelCase = self.query(self.right(_A ) , mid + 1 , _A , _A , _A ) return max(_A , _A ) def __str__( self : Tuple ): return str([self.query(1 , 1 , self.size , _A , _A ) for i in range(1 , self.size + 1 )] ) if __name__ == "__main__": _lowerCAmelCase = [1, 2, -4, 7, 3, -5, 6, 11, -20, 9, 14, 15, 5, 2, -8] _lowerCAmelCase = 15 _lowerCAmelCase = SegmentTree(size) segt.build(1, 1, size, A) print(segt.query(1, 1, size, 4, 6)) print(segt.query(1, 1, size, 7, 11)) print(segt.query(1, 1, size, 7, 12)) segt.update(1, 1, size, 1, 3, 111) print(segt.query(1, 1, size, 1, 15)) segt.update(1, 1, size, 7, 8, 235) print(segt)
71
1
def _snake_case ( __snake_case ): _UpperCamelCase = len(__snake_case ) for i in range(length - 1 ): _UpperCamelCase = i for k in range(i + 1 , __snake_case ): if collection[k] < collection[least]: _UpperCamelCase = k if least != i: _UpperCamelCase , _UpperCamelCase = (collection[i], collection[least]) return collection if __name__ == "__main__": _lowerCAmelCase = input("Enter numbers separated by a comma:\n").strip() _lowerCAmelCase = [int(item) for item in user_input.split(",")] print(selection_sort(unsorted))
71
from typing import TYPE_CHECKING from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available _lowerCAmelCase = { "configuration_jukebox": [ "JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP", "JukeboxConfig", "JukeboxPriorConfig", "JukeboxVQVAEConfig", ], "tokenization_jukebox": ["JukeboxTokenizer"], } try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST", "JukeboxModel", "JukeboxPreTrainedModel", "JukeboxVQVAE", "JukeboxPrior", ] if TYPE_CHECKING: from .configuration_jukebox import ( JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP, JukeboxConfig, JukeboxPriorConfig, JukeboxVQVAEConfig, ) from .tokenization_jukebox import JukeboxTokenizer try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_jukebox import ( JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST, JukeboxModel, JukeboxPreTrainedModel, JukeboxPrior, JukeboxVQVAE, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
1
import json import os import unittest from transformers import BatchEncoding, MvpTokenizer, MvpTokenizerFast from transformers.models.roberta.tokenization_roberta import VOCAB_FILES_NAMES from transformers.testing_utils import require_tokenizers, require_torch from transformers.utils import cached_property from ...test_tokenization_common import TokenizerTesterMixin, filter_roberta_detectors @require_tokenizers class lowerCAmelCase_ ( __lowercase, unittest.TestCase ): UpperCAmelCase = MvpTokenizer UpperCAmelCase = MvpTokenizerFast UpperCAmelCase = True UpperCAmelCase = filter_roberta_detectors def UpperCamelCase_ ( self : Union[str, Any] ): super().setUp() _UpperCamelCase = [ '''l''', '''o''', '''w''', '''e''', '''r''', '''s''', '''t''', '''i''', '''d''', '''n''', '''\u0120''', '''\u0120l''', '''\u0120n''', '''\u0120lo''', '''\u0120low''', '''er''', '''\u0120lowest''', '''\u0120newer''', '''\u0120wider''', '''<unk>''', ] _UpperCamelCase = dict(zip(_A , range(len(_A ) ) ) ) _UpperCamelCase = ['''#version: 0.2''', '''\u0120 l''', '''\u0120l o''', '''\u0120lo w''', '''e r''', ''''''] _UpperCamelCase = {'''unk_token''': '''<unk>'''} _UpperCamelCase = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES['''vocab_file'''] ) _UpperCamelCase = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES['''merges_file'''] ) with open(self.vocab_file , '''w''' , encoding='''utf-8''' ) as fp: fp.write(json.dumps(_A ) + '''\n''' ) with open(self.merges_file , '''w''' , encoding='''utf-8''' ) as fp: fp.write('''\n'''.join(_A ) ) def UpperCamelCase_ ( self : int , **_A : Any ): kwargs.update(self.special_tokens_map ) return self.tokenizer_class.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : List[str] , **_A : Tuple ): kwargs.update(self.special_tokens_map ) return self.rust_tokenizer_class.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : Optional[int] , _A : Any ): return "lower newer", "lower newer" @cached_property def UpperCamelCase_ ( self : Optional[int] ): return MvpTokenizer.from_pretrained('''RUCAIBox/mvp''' ) @cached_property def UpperCamelCase_ ( self : Any ): return MvpTokenizerFast.from_pretrained('''RUCAIBox/mvp''' ) @require_torch def UpperCamelCase_ ( self : Any ): _UpperCamelCase = ['''A long paragraph for summarization.''', '''Another paragraph for summarization.'''] _UpperCamelCase = [0, 250, 251, 1_7818, 13, 3_9186, 1938, 4, 2] for tokenizer in [self.default_tokenizer, self.default_tokenizer_fast]: _UpperCamelCase = tokenizer(_A , max_length=len(_A ) , padding=_A , return_tensors='''pt''' ) self.assertIsInstance(_A , _A ) self.assertEqual((2, 9) , batch.input_ids.shape ) self.assertEqual((2, 9) , batch.attention_mask.shape ) _UpperCamelCase = batch.input_ids.tolist()[0] self.assertListEqual(_A , _A ) # Test that special tokens are reset @require_torch def UpperCamelCase_ ( self : str ): _UpperCamelCase = ['''A long paragraph for summarization.''', '''Another paragraph for summarization.'''] for tokenizer in [self.default_tokenizer, self.default_tokenizer_fast]: _UpperCamelCase = tokenizer(_A , padding=_A , return_tensors='''pt''' ) # check if input_ids are returned and no labels self.assertIn('''input_ids''' , _A ) self.assertIn('''attention_mask''' , _A ) self.assertNotIn('''labels''' , _A ) self.assertNotIn('''decoder_attention_mask''' , _A ) @require_torch def UpperCamelCase_ ( self : int ): _UpperCamelCase = [ '''Summary of the text.''', '''Another summary.''', ] for tokenizer in [self.default_tokenizer, self.default_tokenizer_fast]: _UpperCamelCase = tokenizer(text_target=_A , max_length=32 , padding='''max_length''' , return_tensors='''pt''' ) self.assertEqual(32 , targets['''input_ids'''].shape[1] ) @require_torch def UpperCamelCase_ ( self : List[str] ): for tokenizer in [self.default_tokenizer, self.default_tokenizer_fast]: _UpperCamelCase = tokenizer( ['''I am a small frog''' * 1024, '''I am a small frog'''] , padding=_A , truncation=_A , return_tensors='''pt''' ) self.assertIsInstance(_A , _A ) self.assertEqual(batch.input_ids.shape , (2, 1024) ) @require_torch def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = ['''A long paragraph for summarization.'''] _UpperCamelCase = [ '''Summary of the text.''', ] for tokenizer in [self.default_tokenizer, self.default_tokenizer_fast]: _UpperCamelCase = tokenizer(_A , text_target=_A , return_tensors='''pt''' ) _UpperCamelCase = inputs['''input_ids'''] _UpperCamelCase = inputs['''labels'''] self.assertTrue((input_ids[:, 0] == tokenizer.bos_token_id).all().item() ) self.assertTrue((labels[:, 0] == tokenizer.bos_token_id).all().item() ) self.assertTrue((input_ids[:, -1] == tokenizer.eos_token_id).all().item() ) self.assertTrue((labels[:, -1] == tokenizer.eos_token_id).all().item() ) def UpperCamelCase_ ( self : Dict ): pass def UpperCamelCase_ ( self : Dict ): for tokenizer, pretrained_name, kwargs in self.tokenizers_list: with self.subTest(F"""{tokenizer.__class__.__name__} ({pretrained_name})""" ): _UpperCamelCase = self.rust_tokenizer_class.from_pretrained(_A , **_A ) _UpperCamelCase = self.tokenizer_class.from_pretrained(_A , **_A ) _UpperCamelCase = '''A, <mask> AllenNLP sentence.''' _UpperCamelCase = tokenizer_r.encode_plus(_A , add_special_tokens=_A , return_token_type_ids=_A ) _UpperCamelCase = tokenizer_p.encode_plus(_A , add_special_tokens=_A , return_token_type_ids=_A ) # token_type_ids should put 0 everywhere self.assertEqual(sum(tokens_r['''token_type_ids'''] ) , sum(tokens_p['''token_type_ids'''] ) ) # attention_mask should put 1 everywhere, so sum over length should be 1 self.assertEqual( sum(tokens_r['''attention_mask'''] ) / len(tokens_r['''attention_mask'''] ) , sum(tokens_p['''attention_mask'''] ) / len(tokens_p['''attention_mask'''] ) , ) _UpperCamelCase = tokenizer_r.convert_ids_to_tokens(tokens_r['''input_ids'''] ) _UpperCamelCase = tokenizer_p.convert_ids_to_tokens(tokens_p['''input_ids'''] ) # Rust correctly handles the space before the mask while python doesnt self.assertSequenceEqual(tokens_p['''input_ids'''] , [0, 250, 6, 5_0264, 3823, 487, 2_1992, 3645, 4, 2] ) self.assertSequenceEqual(tokens_r['''input_ids'''] , [0, 250, 6, 5_0264, 3823, 487, 2_1992, 3645, 4, 2] ) self.assertSequenceEqual( _A , ['''<s>''', '''A''', ''',''', '''<mask>''', '''ĠAllen''', '''N''', '''LP''', '''Ġsentence''', '''.''', '''</s>'''] ) self.assertSequenceEqual( _A , ['''<s>''', '''A''', ''',''', '''<mask>''', '''ĠAllen''', '''N''', '''LP''', '''Ġsentence''', '''.''', '''</s>'''] )
71
import multiprocessing import os from typing import BinaryIO, Optional, Union import fsspec from .. import Dataset, Features, NamedSplit, config from ..formatting import query_table from ..packaged_modules.json.json import Json from ..utils import logging from ..utils.typing import NestedDataStructureLike, PathLike from .abc import AbstractDatasetReader class lowerCAmelCase_ ( __lowercase ): def __init__( self : int , _A : NestedDataStructureLike[PathLike] , _A : Optional[NamedSplit] = None , _A : Optional[Features] = None , _A : str = None , _A : bool = False , _A : bool = False , _A : Optional[str] = None , _A : Optional[int] = None , **_A : str , ): super().__init__( _A , split=_A , features=_A , cache_dir=_A , keep_in_memory=_A , streaming=_A , num_proc=_A , **_A , ) _UpperCamelCase = field _UpperCamelCase = path_or_paths if isinstance(_A , _A ) else {self.split: path_or_paths} _UpperCamelCase = Json( cache_dir=_A , data_files=_A , features=_A , field=_A , **_A , ) def UpperCamelCase_ ( self : List[str] ): # Build iterable dataset if self.streaming: _UpperCamelCase = self.builder.as_streaming_dataset(split=self.split ) # Build regular (map-style) dataset else: _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None self.builder.download_and_prepare( download_config=_A , download_mode=_A , verification_mode=_A , base_path=_A , num_proc=self.num_proc , ) _UpperCamelCase = self.builder.as_dataset( split=self.split , verification_mode=_A , in_memory=self.keep_in_memory ) return dataset class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : Dataset , _A : Union[PathLike, BinaryIO] , _A : Optional[int] = None , _A : Optional[int] = None , **_A : List[str] , ): if num_proc is not None and num_proc <= 0: raise ValueError(F"""num_proc {num_proc} must be an integer > 0.""" ) _UpperCamelCase = dataset _UpperCamelCase = path_or_buf _UpperCamelCase = batch_size if batch_size else config.DEFAULT_MAX_BATCH_SIZE _UpperCamelCase = num_proc _UpperCamelCase = '''utf-8''' _UpperCamelCase = to_json_kwargs def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = self.to_json_kwargs.pop('''path_or_buf''' , _A ) _UpperCamelCase = self.to_json_kwargs.pop('''orient''' , '''records''' ) _UpperCamelCase = self.to_json_kwargs.pop('''lines''' , True if orient == '''records''' else False ) _UpperCamelCase = self.to_json_kwargs.pop('''index''' , False if orient in ['''split''', '''table'''] else True ) _UpperCamelCase = self.to_json_kwargs.pop('''compression''' , _A ) if compression not in [None, "infer", "gzip", "bz2", "xz"]: raise NotImplementedError(F"""`datasets` currently does not support {compression} compression""" ) if isinstance(self.path_or_buf , (str, bytes, os.PathLike) ): with fsspec.open(self.path_or_buf , '''wb''' , compression=_A ) as buffer: _UpperCamelCase = self._write(file_obj=_A , orient=_A , lines=_A , index=_A , **self.to_json_kwargs ) else: if compression: raise NotImplementedError( F"""The compression parameter is not supported when writing to a buffer, but compression={compression}""" ''' was passed. Please provide a local path instead.''' ) _UpperCamelCase = self._write( file_obj=self.path_or_buf , orient=_A , lines=_A , index=_A , **self.to_json_kwargs ) return written def UpperCamelCase_ ( self : Any , _A : Optional[Any] ): _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = args _UpperCamelCase = query_table( table=self.dataset.data , key=slice(_A , offset + self.batch_size ) , indices=self.dataset._indices , ) _UpperCamelCase = batch.to_pandas().to_json( path_or_buf=_A , orient=_A , lines=_A , index=_A , **_A ) if not json_str.endswith('''\n''' ): json_str += "\n" return json_str.encode(self.encoding ) def UpperCamelCase_ ( self : int , _A : BinaryIO , _A : Dict , _A : Optional[Any] , _A : Dict , **_A : str , ): _UpperCamelCase = 0 if self.num_proc is None or self.num_proc == 1: for offset in logging.tqdm( range(0 , len(self.dataset ) , self.batch_size ) , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating json from Arrow format''' , ): _UpperCamelCase = self._batch_json((offset, orient, lines, index, to_json_kwargs) ) written += file_obj.write(_A ) else: _UpperCamelCase , _UpperCamelCase = len(self.dataset ), self.batch_size with multiprocessing.Pool(self.num_proc ) as pool: for json_str in logging.tqdm( pool.imap( self._batch_json , [(offset, orient, lines, index, to_json_kwargs) for offset in range(0 , _A , _A )] , ) , total=(num_rows // batch_size) + 1 if num_rows % batch_size else num_rows // batch_size , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating json from Arrow format''' , ): written += file_obj.write(_A ) return written
71
1
from typing import Dict, List from nltk.translate import gleu_score import datasets from datasets import MetricInfo _lowerCAmelCase = "\\n@misc{wu2016googles,\n title={Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation},\n author={Yonghui Wu and Mike Schuster and Zhifeng Chen and Quoc V. Le and Mohammad Norouzi and Wolfgang Macherey\n and Maxim Krikun and Yuan Cao and Qin Gao and Klaus Macherey and Jeff Klingner and Apurva Shah and Melvin\n Johnson and Xiaobing Liu and Łukasz Kaiser and Stephan Gouws and Yoshikiyo Kato and Taku Kudo and Hideto\n Kazawa and Keith Stevens and George Kurian and Nishant Patil and Wei Wang and Cliff Young and\n Jason Smith and Jason Riesa and Alex Rudnick and Oriol Vinyals and Greg Corrado and Macduff Hughes\n and Jeffrey Dean},\n year={2016},\n eprint={1609.08144},\n archivePrefix={arXiv},\n primaryClass={cs.CL}\n}\n" _lowerCAmelCase = "\\nThe BLEU score has some undesirable properties when used for single\nsentences, as it was designed to be a corpus measure. We therefore\nuse a slightly different score for our RL experiments which we call\nthe 'GLEU score'. For the GLEU score, we record all sub-sequences of\n1, 2, 3 or 4 tokens in output and target sequence (n-grams). We then\ncompute a recall, which is the ratio of the number of matching n-grams\nto the number of total n-grams in the target (ground truth) sequence,\nand a precision, which is the ratio of the number of matching n-grams\nto the number of total n-grams in the generated output sequence. Then\nGLEU score is simply the minimum of recall and precision. This GLEU\nscore's range is always between 0 (no matches) and 1 (all match) and\nit is symmetrical when switching output and target. According to\nour experiments, GLEU score correlates quite well with the BLEU\nmetric on a corpus level but does not have its drawbacks for our per\nsentence reward objective.\n" _lowerCAmelCase = "\\nComputes corpus-level Google BLEU (GLEU) score of translated segments against one or more references.\nInstead of averaging the sentence level GLEU scores (i.e. macro-average precision), Wu et al. (2016) sum up the matching\ntokens and the max of hypothesis and reference tokens for each sentence, then compute using the aggregate values.\n\nArgs:\n predictions (list of str): list of translations to score.\n Each translation should be tokenized into a list of tokens.\n references (list of list of str): list of lists of references for each translation.\n Each reference should be tokenized into a list of tokens.\n min_len (int): The minimum order of n-gram this function should extract. Defaults to 1.\n max_len (int): The maximum order of n-gram this function should extract. Defaults to 4.\n\nReturns:\n 'google_bleu': google_bleu score\n\nExamples:\n Example 1:\n >>> hyp1 = ['It', 'is', 'a', 'guide', 'to', 'action', 'which',\n ... 'ensures', 'that', 'the', 'rubber', 'duck', 'always',\n ... 'disobeys', 'the', 'commands', 'of', 'the', 'cat']\n >>> ref1a = ['It', 'is', 'the', 'guiding', 'principle', 'which',\n ... 'guarantees', 'the', 'rubber', 'duck', 'forces', 'never',\n ... 'being', 'under', 'the', 'command', 'of', 'the', 'cat']\n\n >>> hyp2 = ['he', 'read', 'the', 'book', 'because', 'he', 'was',\n ... 'interested', 'in', 'world', 'history']\n >>> ref2a = ['he', 'was', 'interested', 'in', 'world', 'history',\n ... 'because', 'he', 'read', 'the', 'book']\n\n >>> list_of_references = [[ref1a], [ref2a]]\n >>> hypotheses = [hyp1, hyp2]\n >>> google_bleu = datasets.load_metric(\"google_bleu\")\n >>> results = google_bleu.compute(predictions=hypotheses, references=list_of_references)\n >>> print(round(results[\"google_bleu\"], 2))\n 0.44\n\n Example 2:\n >>> hyp1 = ['It', 'is', 'a', 'guide', 'to', 'action', 'which',\n ... 'ensures', 'that', 'the', 'rubber', 'duck', 'always',\n ... 'disobeys', 'the', 'commands', 'of', 'the', 'cat']\n >>> ref1a = ['It', 'is', 'the', 'guiding', 'principle', 'which',\n ... 'guarantees', 'the', 'rubber', 'duck', 'forces', 'never',\n ... 'being', 'under', 'the', 'command', 'of', 'the', 'cat']\n >>> ref1b = ['It', 'is', 'a', 'guide', 'to', 'action', 'that',\n ... 'ensures', 'that', 'the', 'rubber', 'duck', 'will', 'never',\n ... 'heed', 'the', 'cat', 'commands']\n >>> ref1c = ['It', 'is', 'the', 'practical', 'guide', 'for', 'the',\n ... 'rubber', 'duck', 'army', 'never', 'to', 'heed', 'the', 'directions',\n ... 'of', 'the', 'cat']\n\n >>> hyp2 = ['he', 'read', 'the', 'book', 'because', 'he', 'was',\n ... 'interested', 'in', 'world', 'history']\n >>> ref2a = ['he', 'was', 'interested', 'in', 'world', 'history',\n ... 'because', 'he', 'read', 'the', 'book']\n\n >>> list_of_references = [[ref1a, ref1b, ref1c], [ref2a]]\n >>> hypotheses = [hyp1, hyp2]\n >>> google_bleu = datasets.load_metric(\"google_bleu\")\n >>> results = google_bleu.compute(predictions=hypotheses, references=list_of_references)\n >>> print(round(results[\"google_bleu\"], 2))\n 0.61\n\n Example 3:\n >>> hyp1 = ['It', 'is', 'a', 'guide', 'to', 'action', 'which',\n ... 'ensures', 'that', 'the', 'rubber', 'duck', 'always',\n ... 'disobeys', 'the', 'commands', 'of', 'the', 'cat']\n >>> ref1a = ['It', 'is', 'the', 'guiding', 'principle', 'which',\n ... 'guarantees', 'the', 'rubber', 'duck', 'forces', 'never',\n ... 'being', 'under', 'the', 'command', 'of', 'the', 'cat']\n >>> ref1b = ['It', 'is', 'a', 'guide', 'to', 'action', 'that',\n ... 'ensures', 'that', 'the', 'rubber', 'duck', 'will', 'never',\n ... 'heed', 'the', 'cat', 'commands']\n >>> ref1c = ['It', 'is', 'the', 'practical', 'guide', 'for', 'the',\n ... 'rubber', 'duck', 'army', 'never', 'to', 'heed', 'the', 'directions',\n ... 'of', 'the', 'cat']\n\n >>> hyp2 = ['he', 'read', 'the', 'book', 'because', 'he', 'was',\n ... 'interested', 'in', 'world', 'history']\n >>> ref2a = ['he', 'was', 'interested', 'in', 'world', 'history',\n ... 'because', 'he', 'read', 'the', 'book']\n\n >>> list_of_references = [[ref1a, ref1b, ref1c], [ref2a]]\n >>> hypotheses = [hyp1, hyp2]\n >>> google_bleu = datasets.load_metric(\"google_bleu\")\n >>> results = google_bleu.compute(predictions=hypotheses, references=list_of_references, min_len=2)\n >>> print(round(results[\"google_bleu\"], 2))\n 0.53\n\n Example 4:\n >>> hyp1 = ['It', 'is', 'a', 'guide', 'to', 'action', 'which',\n ... 'ensures', 'that', 'the', 'rubber', 'duck', 'always',\n ... 'disobeys', 'the', 'commands', 'of', 'the', 'cat']\n >>> ref1a = ['It', 'is', 'the', 'guiding', 'principle', 'which',\n ... 'guarantees', 'the', 'rubber', 'duck', 'forces', 'never',\n ... 'being', 'under', 'the', 'command', 'of', 'the', 'cat']\n >>> ref1b = ['It', 'is', 'a', 'guide', 'to', 'action', 'that',\n ... 'ensures', 'that', 'the', 'rubber', 'duck', 'will', 'never',\n ... 'heed', 'the', 'cat', 'commands']\n >>> ref1c = ['It', 'is', 'the', 'practical', 'guide', 'for', 'the',\n ... 'rubber', 'duck', 'army', 'never', 'to', 'heed', 'the', 'directions',\n ... 'of', 'the', 'cat']\n\n >>> hyp2 = ['he', 'read', 'the', 'book', 'because', 'he', 'was',\n ... 'interested', 'in', 'world', 'history']\n >>> ref2a = ['he', 'was', 'interested', 'in', 'world', 'history',\n ... 'because', 'he', 'read', 'the', 'book']\n\n >>> list_of_references = [[ref1a, ref1b, ref1c], [ref2a]]\n >>> hypotheses = [hyp1, hyp2]\n >>> google_bleu = datasets.load_metric(\"google_bleu\")\n >>> results = google_bleu.compute(predictions=hypotheses,references=list_of_references, min_len=2, max_len=6)\n >>> print(round(results[\"google_bleu\"], 2))\n 0.4\n" @datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION ) class lowerCAmelCase_ ( datasets.Metric ): def UpperCamelCase_ ( self : Optional[Any] ): return datasets.MetricInfo( description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features( { '''predictions''': datasets.Sequence(datasets.Value('''string''' , id='''token''' ) , id='''sequence''' ), '''references''': datasets.Sequence( datasets.Sequence(datasets.Value('''string''' , id='''token''' ) , id='''sequence''' ) , id='''references''' ), } ) , ) def UpperCamelCase_ ( self : int , _A : List[List[List[str]]] , _A : List[List[str]] , _A : int = 1 , _A : int = 4 , ): return { "google_bleu": gleu_score.corpus_gleu( list_of_references=_A , hypotheses=_A , min_len=_A , max_len=_A ) }
71
import enum import warnings from .. import MODEL_FOR_CAUSAL_LM_MAPPING, TF_MODEL_FOR_CAUSAL_LM_MAPPING from ..utils import add_end_docstrings, is_tf_available from .base import PIPELINE_INIT_ARGS, Pipeline if is_tf_available(): import tensorflow as tf class lowerCAmelCase_ ( enum.Enum ): UpperCAmelCase = 0 UpperCAmelCase = 1 UpperCAmelCase = 2 @add_end_docstrings(__lowercase ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "\n In 1991, the remains of Russian Tsar Nicholas II and his family (except for Alexei and Maria) are discovered. The\n voice of Nicholas's young son, Tsarevich Alexei Nikolaevich, narrates the remainder of the story. 1883 Western\n Siberia, a young Grigori Rasputin is asked by his father and a group of men to perform magic. Rasputin has a vision\n and denounces one of the men as a horse thief. Although his father initially slaps him for making such an\n accusation, Rasputin watches as the man is chased outside and beaten. Twenty years later, Rasputin sees a vision of\n the Virgin Mary, prompting him to become a priest. Rasputin quickly becomes famous, with people, even a bishop,\n begging for his blessing. <eod> </s> <eos>\n " def __init__( self : Tuple , *_A : List[str] , **_A : str ): super().__init__(*_A , **_A ) self.check_model_type( TF_MODEL_FOR_CAUSAL_LM_MAPPING if self.framework == '''tf''' else MODEL_FOR_CAUSAL_LM_MAPPING ) if "prefix" not in self._preprocess_params: # This is very specific. The logic is quite complex and needs to be done # as a "default". # It also defines both some preprocess_kwargs and generate_kwargs # which is why we cannot put them in their respective methods. _UpperCamelCase = None if self.model.config.prefix is not None: _UpperCamelCase = self.model.config.prefix if prefix is None and self.model.__class__.__name__ in [ "XLNetLMHeadModel", "TransfoXLLMHeadModel", "TFXLNetLMHeadModel", "TFTransfoXLLMHeadModel", ]: # For XLNet and TransformerXL we add an article to the prompt to give more state to the model. _UpperCamelCase = self.XL_PREFIX if prefix is not None: # Recalculate some generate_kwargs linked to prefix. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = self._sanitize_parameters(prefix=_A , **self._forward_params ) _UpperCamelCase = {**self._preprocess_params, **preprocess_params} _UpperCamelCase = {**self._forward_params, **forward_params} def UpperCamelCase_ ( self : Dict , _A : Optional[int]=None , _A : Any=None , _A : Optional[int]=None , _A : List[str]=None , _A : List[Any]=None , _A : int=None , _A : Tuple=None , _A : Optional[Any]=None , **_A : Optional[int] , ): _UpperCamelCase = {} if prefix is not None: _UpperCamelCase = prefix if prefix: _UpperCamelCase = self.tokenizer( _A , padding=_A , add_special_tokens=_A , return_tensors=self.framework ) _UpperCamelCase = prefix_inputs['''input_ids'''].shape[-1] if handle_long_generation is not None: if handle_long_generation not in {"hole"}: raise ValueError( F"""{handle_long_generation} is not a valid value for `handle_long_generation` parameter expected""" ''' [None, \'hole\']''' ) _UpperCamelCase = handle_long_generation preprocess_params.update(_A ) _UpperCamelCase = generate_kwargs _UpperCamelCase = {} if return_full_text is not None and return_type is None: if return_text is not None: raise ValueError('''`return_text` is mutually exclusive with `return_full_text`''' ) if return_tensors is not None: raise ValueError('''`return_full_text` is mutually exclusive with `return_tensors`''' ) _UpperCamelCase = ReturnType.FULL_TEXT if return_full_text else ReturnType.NEW_TEXT if return_tensors is not None and return_type is None: if return_text is not None: raise ValueError('''`return_text` is mutually exclusive with `return_tensors`''' ) _UpperCamelCase = ReturnType.TENSORS if return_type is not None: _UpperCamelCase = return_type if clean_up_tokenization_spaces is not None: _UpperCamelCase = clean_up_tokenization_spaces if stop_sequence is not None: _UpperCamelCase = self.tokenizer.encode(_A , add_special_tokens=_A ) if len(_A ) > 1: warnings.warn( '''Stopping on a multiple token sequence is not yet supported on transformers. The first token of''' ''' the stop sequence will be used as the stop sequence string in the interim.''' ) _UpperCamelCase = stop_sequence_ids[0] return preprocess_params, forward_params, postprocess_params def UpperCamelCase_ ( self : int , *_A : Union[str, Any] , **_A : Union[str, Any] ): # Parse arguments if self.model.__class__.__name__ in ["TransfoXLLMHeadModel"]: kwargs.update({'''add_space_before_punct_symbol''': True} ) return super()._parse_and_tokenize(*_A , **_A ) def __call__( self : List[str] , _A : str , **_A : Any ): return super().__call__(_A , **_A ) def UpperCamelCase_ ( self : Optional[Any] , _A : List[str] , _A : int="" , _A : Optional[Any]=None , **_A : Optional[Any] ): _UpperCamelCase = self.tokenizer( prefix + prompt_text , padding=_A , add_special_tokens=_A , return_tensors=self.framework ) _UpperCamelCase = prompt_text if handle_long_generation == "hole": _UpperCamelCase = inputs['''input_ids'''].shape[-1] if "max_new_tokens" in generate_kwargs: _UpperCamelCase = generate_kwargs['''max_new_tokens'''] else: _UpperCamelCase = generate_kwargs.get('''max_length''' , self.model.config.max_length ) - cur_len if new_tokens < 0: raise ValueError('''We cannot infer how many new tokens are expected''' ) if cur_len + new_tokens > self.tokenizer.model_max_length: _UpperCamelCase = self.tokenizer.model_max_length - new_tokens if keep_length <= 0: raise ValueError( '''We cannot use `hole` to handle this generation the number of desired tokens exceeds the''' ''' models max length''' ) _UpperCamelCase = inputs['''input_ids'''][:, -keep_length:] if "attention_mask" in inputs: _UpperCamelCase = inputs['''attention_mask'''][:, -keep_length:] return inputs def UpperCamelCase_ ( self : Dict , _A : Optional[int] , **_A : str ): _UpperCamelCase = model_inputs['''input_ids'''] _UpperCamelCase = model_inputs.get('''attention_mask''' , _A ) # Allow empty prompts if input_ids.shape[1] == 0: _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = 1 else: _UpperCamelCase = input_ids.shape[0] _UpperCamelCase = model_inputs.pop('''prompt_text''' ) # If there is a prefix, we may need to adjust the generation length. Do so without permanently modifying # generate_kwargs, as some of the parameterization may come from the initialization of the pipeline. _UpperCamelCase = generate_kwargs.pop('''prefix_length''' , 0 ) if prefix_length > 0: _UpperCamelCase = '''max_new_tokens''' in generate_kwargs or ( '''generation_config''' in generate_kwargs and generate_kwargs['''generation_config'''].max_new_tokens is not None ) if not has_max_new_tokens: _UpperCamelCase = generate_kwargs.get('''max_length''' ) or self.model.config.max_length generate_kwargs["max_length"] += prefix_length _UpperCamelCase = '''min_new_tokens''' in generate_kwargs or ( '''generation_config''' in generate_kwargs and generate_kwargs['''generation_config'''].min_new_tokens is not None ) if not has_min_new_tokens and "min_length" in generate_kwargs: generate_kwargs["min_length"] += prefix_length # BS x SL _UpperCamelCase = self.model.generate(input_ids=_A , attention_mask=_A , **_A ) _UpperCamelCase = generated_sequence.shape[0] if self.framework == "pt": _UpperCamelCase = generated_sequence.reshape(_A , out_b // in_b , *generated_sequence.shape[1:] ) elif self.framework == "tf": _UpperCamelCase = tf.reshape(_A , (in_b, out_b // in_b, *generated_sequence.shape[1:]) ) return {"generated_sequence": generated_sequence, "input_ids": input_ids, "prompt_text": prompt_text} def UpperCamelCase_ ( self : List[str] , _A : Dict , _A : Optional[Any]=ReturnType.FULL_TEXT , _A : Dict=True ): _UpperCamelCase = model_outputs['''generated_sequence'''][0] _UpperCamelCase = model_outputs['''input_ids'''] _UpperCamelCase = model_outputs['''prompt_text'''] _UpperCamelCase = generated_sequence.numpy().tolist() _UpperCamelCase = [] for sequence in generated_sequence: if return_type == ReturnType.TENSORS: _UpperCamelCase = {'''generated_token_ids''': sequence} elif return_type in {ReturnType.NEW_TEXT, ReturnType.FULL_TEXT}: # Decode text _UpperCamelCase = self.tokenizer.decode( _A , skip_special_tokens=_A , clean_up_tokenization_spaces=_A , ) # Remove PADDING prompt of the sequence if XLNet or Transfo-XL model is used if input_ids is None: _UpperCamelCase = 0 else: _UpperCamelCase = len( self.tokenizer.decode( input_ids[0] , skip_special_tokens=_A , clean_up_tokenization_spaces=_A , ) ) if return_type == ReturnType.FULL_TEXT: _UpperCamelCase = prompt_text + text[prompt_length:] else: _UpperCamelCase = text[prompt_length:] _UpperCamelCase = {'''generated_text''': all_text} records.append(_A ) return records
71
1
import inspect import unittest import warnings from transformers import DeiTConfig from transformers.models.auto import get_values from transformers.testing_utils import ( require_accelerate, require_torch, require_torch_gpu, require_vision, slow, torch_device, ) from transformers.utils import cached_property, is_torch_available, is_vision_available from ...test_configuration_common import ConfigTester from ...test_modeling_common import ModelTesterMixin, floats_tensor, ids_tensor from ...test_pipeline_mixin import PipelineTesterMixin if is_torch_available(): import torch from torch import nn from transformers import ( MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING, MODEL_FOR_SEQUENCE_CLASSIFICATION_MAPPING, MODEL_MAPPING, DeiTForImageClassification, DeiTForImageClassificationWithTeacher, DeiTForMaskedImageModeling, DeiTModel, ) from transformers.models.deit.modeling_deit import DEIT_PRETRAINED_MODEL_ARCHIVE_LIST if is_vision_available(): from PIL import Image from transformers import DeiTImageProcessor class lowerCAmelCase_ : def __init__( self : Union[str, Any] , _A : Optional[int] , _A : Optional[Any]=13 , _A : List[Any]=30 , _A : Any=2 , _A : Optional[int]=3 , _A : Optional[Any]=True , _A : Dict=True , _A : Optional[Any]=32 , _A : Optional[Any]=5 , _A : List[str]=4 , _A : Tuple=37 , _A : Optional[int]="gelu" , _A : int=0.1 , _A : List[Any]=0.1 , _A : int=10 , _A : Any=0.02 , _A : int=3 , _A : Optional[Any]=None , _A : Optional[int]=2 , ): _UpperCamelCase = parent _UpperCamelCase = batch_size _UpperCamelCase = image_size _UpperCamelCase = patch_size _UpperCamelCase = num_channels _UpperCamelCase = is_training _UpperCamelCase = use_labels _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = hidden_dropout_prob _UpperCamelCase = attention_probs_dropout_prob _UpperCamelCase = type_sequence_label_size _UpperCamelCase = initializer_range _UpperCamelCase = scope _UpperCamelCase = encoder_stride # in DeiT, the seq length equals the number of patches + 2 (we add 2 for the [CLS] and distilation tokens) _UpperCamelCase = (image_size // patch_size) ** 2 _UpperCamelCase = num_patches + 2 def UpperCamelCase_ ( self : Any ): _UpperCamelCase = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] ) _UpperCamelCase = None if self.use_labels: _UpperCamelCase = ids_tensor([self.batch_size] , self.type_sequence_label_size ) _UpperCamelCase = self.get_config() return config, pixel_values, labels def UpperCamelCase_ ( self : Tuple ): return DeiTConfig( image_size=self.image_size , patch_size=self.patch_size , num_channels=self.num_channels , hidden_size=self.hidden_size , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , is_decoder=_A , initializer_range=self.initializer_range , encoder_stride=self.encoder_stride , ) def UpperCamelCase_ ( self : str , _A : List[str] , _A : str , _A : Tuple ): _UpperCamelCase = DeiTModel(config=_A ) model.to(_A ) model.eval() _UpperCamelCase = model(_A ) self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) ) def UpperCamelCase_ ( self : Optional[int] , _A : Dict , _A : Optional[Any] , _A : Any ): _UpperCamelCase = DeiTForMaskedImageModeling(config=_A ) model.to(_A ) model.eval() _UpperCamelCase = model(_A ) self.parent.assertEqual( result.reconstruction.shape , (self.batch_size, self.num_channels, self.image_size, self.image_size) ) # test greyscale images _UpperCamelCase = 1 _UpperCamelCase = DeiTForMaskedImageModeling(_A ) model.to(_A ) model.eval() _UpperCamelCase = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] ) _UpperCamelCase = model(_A ) self.parent.assertEqual(result.reconstruction.shape , (self.batch_size, 1, self.image_size, self.image_size) ) def UpperCamelCase_ ( self : int , _A : Dict , _A : Optional[Any] , _A : int ): _UpperCamelCase = self.type_sequence_label_size _UpperCamelCase = DeiTForImageClassification(_A ) model.to(_A ) model.eval() _UpperCamelCase = model(_A , labels=_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) ) # test greyscale images _UpperCamelCase = 1 _UpperCamelCase = DeiTForImageClassification(_A ) model.to(_A ) model.eval() _UpperCamelCase = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] ) _UpperCamelCase = model(_A , labels=_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.prepare_config_and_inputs() ( ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ) = config_and_inputs _UpperCamelCase = {'''pixel_values''': pixel_values} return config, inputs_dict @require_torch class lowerCAmelCase_ ( __lowercase, __lowercase, unittest.TestCase ): UpperCAmelCase = ( ( DeiTModel, DeiTForImageClassification, DeiTForImageClassificationWithTeacher, DeiTForMaskedImageModeling, ) if is_torch_available() else () ) UpperCAmelCase = ( { "feature-extraction": DeiTModel, "image-classification": (DeiTForImageClassification, DeiTForImageClassificationWithTeacher), } if is_torch_available() else {} ) UpperCAmelCase = False UpperCAmelCase = False UpperCAmelCase = False def UpperCamelCase_ ( self : int ): _UpperCamelCase = DeiTModelTester(self ) _UpperCamelCase = ConfigTester(self , config_class=_A , has_text_modality=_A , hidden_size=37 ) def UpperCamelCase_ ( self : Dict ): self.config_tester.run_common_tests() @unittest.skip(reason='''DeiT does not use inputs_embeds''' ) def UpperCamelCase_ ( self : List[Any] ): pass def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase , _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() for model_class in self.all_model_classes: _UpperCamelCase = model_class(_A ) self.assertIsInstance(model.get_input_embeddings() , (nn.Module) ) _UpperCamelCase = model.get_output_embeddings() self.assertTrue(x is None or isinstance(_A , nn.Linear ) ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase , _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() for model_class in self.all_model_classes: _UpperCamelCase = model_class(_A ) _UpperCamelCase = inspect.signature(model.forward ) # signature.parameters is an OrderedDict => so arg_names order is deterministic _UpperCamelCase = [*signature.parameters.keys()] _UpperCamelCase = ['''pixel_values'''] self.assertListEqual(arg_names[:1] , _A ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model(*_A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_masked_image_modeling(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_image_classification(*_A ) def UpperCamelCase_ ( self : Optional[Any] , _A : int , _A : List[Any] , _A : Any=False ): _UpperCamelCase = super()._prepare_for_class(_A , _A , return_labels=_A ) if return_labels: if model_class.__name__ == "DeiTForImageClassificationWithTeacher": del inputs_dict["labels"] return inputs_dict def UpperCamelCase_ ( self : str ): if not self.model_tester.is_training: return _UpperCamelCase , _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() _UpperCamelCase = True for model_class in self.all_model_classes: # DeiTForImageClassificationWithTeacher supports inference-only if ( model_class in get_values(_A ) or model_class.__name__ == "DeiTForImageClassificationWithTeacher" ): continue _UpperCamelCase = model_class(_A ) model.to(_A ) model.train() _UpperCamelCase = self._prepare_for_class(_A , _A , return_labels=_A ) _UpperCamelCase = model(**_A ).loss loss.backward() def UpperCamelCase_ ( self : str ): _UpperCamelCase , _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() if not self.model_tester.is_training: return _UpperCamelCase = False _UpperCamelCase = True for model_class in self.all_model_classes: if model_class in get_values(_A ) or not model_class.supports_gradient_checkpointing: continue # DeiTForImageClassificationWithTeacher supports inference-only if model_class.__name__ == "DeiTForImageClassificationWithTeacher": continue _UpperCamelCase = model_class(_A ) model.gradient_checkpointing_enable() model.to(_A ) model.train() _UpperCamelCase = self._prepare_for_class(_A , _A , return_labels=_A ) _UpperCamelCase = model(**_A ).loss loss.backward() def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase , _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() _UpperCamelCase = [ {'''title''': '''multi_label_classification''', '''num_labels''': 2, '''dtype''': torch.float}, {'''title''': '''single_label_classification''', '''num_labels''': 1, '''dtype''': torch.long}, {'''title''': '''regression''', '''num_labels''': 1, '''dtype''': torch.float}, ] for model_class in self.all_model_classes: if ( model_class not in [ *get_values(_A ), *get_values(_A ), ] or model_class.__name__ == "DeiTForImageClassificationWithTeacher" ): continue for problem_type in problem_types: with self.subTest(msg=F"""Testing {model_class} with {problem_type["title"]}""" ): _UpperCamelCase = problem_type['''title'''] _UpperCamelCase = problem_type['''num_labels'''] _UpperCamelCase = model_class(_A ) model.to(_A ) model.train() _UpperCamelCase = self._prepare_for_class(_A , _A , return_labels=_A ) if problem_type["num_labels"] > 1: _UpperCamelCase = inputs['''labels'''].unsqueeze(1 ).repeat(1 , problem_type['''num_labels'''] ) _UpperCamelCase = inputs['''labels'''].to(problem_type['''dtype'''] ) # This tests that we do not trigger the warning form PyTorch "Using a target size that is different # to the input size. This will likely lead to incorrect results due to broadcasting. Please ensure # they have the same size." which is a symptom something in wrong for the regression problem. # See https://github.com/huggingface/transformers/issues/11780 with warnings.catch_warnings(record=_A ) as warning_list: _UpperCamelCase = model(**_A ).loss for w in warning_list: if "Using a target size that is different to the input size" in str(w.message ): raise ValueError( F"""Something is going wrong in the regression problem: intercepted {w.message}""" ) loss.backward() @slow def UpperCamelCase_ ( self : List[Any] ): for model_name in DEIT_PRETRAINED_MODEL_ARCHIVE_LIST[:1]: _UpperCamelCase = DeiTModel.from_pretrained(_A ) self.assertIsNotNone(_A ) def _snake_case ( ): _UpperCamelCase = Image.open('''./tests/fixtures/tests_samples/COCO/000000039769.png''' ) return image @require_torch @require_vision class lowerCAmelCase_ ( unittest.TestCase ): @cached_property def UpperCamelCase_ ( self : int ): return ( DeiTImageProcessor.from_pretrained('''facebook/deit-base-distilled-patch16-224''' ) if is_vision_available() else None ) @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = DeiTForImageClassificationWithTeacher.from_pretrained('''facebook/deit-base-distilled-patch16-224''' ).to( _A ) _UpperCamelCase = self.default_image_processor _UpperCamelCase = prepare_img() _UpperCamelCase = image_processor(images=_A , return_tensors='''pt''' ).to(_A ) # forward pass with torch.no_grad(): _UpperCamelCase = model(**_A ) # verify the logits _UpperCamelCase = torch.Size((1, 1000) ) self.assertEqual(outputs.logits.shape , _A ) _UpperCamelCase = torch.tensor([-1.0266, 0.1912, -1.2861] ).to(_A ) self.assertTrue(torch.allclose(outputs.logits[0, :3] , _A , atol=1e-4 ) ) @slow @require_accelerate @require_torch_gpu def UpperCamelCase_ ( self : int ): _UpperCamelCase = DeiTModel.from_pretrained( '''facebook/deit-base-distilled-patch16-224''' , torch_dtype=torch.floataa , device_map='''auto''' ) _UpperCamelCase = self.default_image_processor _UpperCamelCase = prepare_img() _UpperCamelCase = image_processor(images=_A , return_tensors='''pt''' ) _UpperCamelCase = inputs.pixel_values.to(_A ) # forward pass to make sure inference works in fp16 with torch.no_grad(): _UpperCamelCase = model(_A )
71
import unittest from transformers import is_torch_available from transformers.testing_utils import require_sentencepiece, require_tokenizers, require_torch, slow, torch_device if is_torch_available(): from transformers import AutoModelForSeqaSeqLM, AutoTokenizer @require_torch @require_sentencepiece @require_tokenizers class lowerCAmelCase_ ( unittest.TestCase ): @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = AutoModelForSeqaSeqLM.from_pretrained('''google/mt5-small''' , return_dict=_A ).to(_A ) _UpperCamelCase = AutoTokenizer.from_pretrained('''google/mt5-small''' ) _UpperCamelCase = tokenizer('''Hello there''' , return_tensors='''pt''' ).input_ids _UpperCamelCase = tokenizer('''Hi I am''' , return_tensors='''pt''' ).input_ids _UpperCamelCase = model(input_ids.to(_A ) , labels=labels.to(_A ) ).loss _UpperCamelCase = -(labels.shape[-1] * loss.item()) _UpperCamelCase = -84.9127 self.assertTrue(abs(mtf_score - EXPECTED_SCORE ) < 1e-4 )
71
1
from typing import Optional, Tuple, Union import flax import flax.linen as nn import jax import jax.numpy as jnp from flax.core.frozen_dict import FrozenDict from ..configuration_utils import ConfigMixin, flax_register_to_config from ..utils import BaseOutput from .embeddings_flax import FlaxTimestepEmbedding, FlaxTimesteps from .modeling_flax_utils import FlaxModelMixin from .unet_ad_blocks_flax import ( FlaxCrossAttnDownBlockaD, FlaxDownBlockaD, FlaxUNetMidBlockaDCrossAttn, ) @flax.struct.dataclass class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = 42 UpperCAmelCase = 42 class lowerCAmelCase_ ( nn.Module ): UpperCAmelCase = 42 UpperCAmelCase = (16, 32, 96, 256) UpperCAmelCase = jnp.floataa def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = nn.Conv( self.block_out_channels[0] , kernel_size=(3, 3) , padding=((1, 1), (1, 1)) , dtype=self.dtype , ) _UpperCamelCase = [] for i in range(len(self.block_out_channels ) - 1 ): _UpperCamelCase = self.block_out_channels[i] _UpperCamelCase = self.block_out_channels[i + 1] _UpperCamelCase = nn.Conv( _A , kernel_size=(3, 3) , padding=((1, 1), (1, 1)) , dtype=self.dtype , ) blocks.append(_A ) _UpperCamelCase = nn.Conv( _A , kernel_size=(3, 3) , strides=(2, 2) , padding=((1, 1), (1, 1)) , dtype=self.dtype , ) blocks.append(_A ) _UpperCamelCase = blocks _UpperCamelCase = nn.Conv( self.conditioning_embedding_channels , kernel_size=(3, 3) , padding=((1, 1), (1, 1)) , kernel_init=nn.initializers.zeros_init() , bias_init=nn.initializers.zeros_init() , dtype=self.dtype , ) def __call__( self : Tuple , _A : Tuple ): _UpperCamelCase = self.conv_in(_A ) _UpperCamelCase = nn.silu(_A ) for block in self.blocks: _UpperCamelCase = block(_A ) _UpperCamelCase = nn.silu(_A ) _UpperCamelCase = self.conv_out(_A ) return embedding @flax_register_to_config class lowerCAmelCase_ ( nn.Module, __lowercase, __lowercase ): UpperCAmelCase = 32 UpperCAmelCase = 4 UpperCAmelCase = ( "CrossAttnDownBlock2D", "CrossAttnDownBlock2D", "CrossAttnDownBlock2D", "DownBlock2D", ) UpperCAmelCase = False UpperCAmelCase = (320, 640, 1280, 1280) UpperCAmelCase = 2 UpperCAmelCase = 8 UpperCAmelCase = None UpperCAmelCase = 1280 UpperCAmelCase = 0.0 UpperCAmelCase = False UpperCAmelCase = jnp.floataa UpperCAmelCase = True UpperCAmelCase = 0 UpperCAmelCase = "rgb" UpperCAmelCase = (16, 32, 96, 256) def UpperCamelCase_ ( self : List[Any] , _A : jax.random.KeyArray ): # init input tensors _UpperCamelCase = (1, self.in_channels, self.sample_size, self.sample_size) _UpperCamelCase = jnp.zeros(_A , dtype=jnp.floataa ) _UpperCamelCase = jnp.ones((1,) , dtype=jnp.intaa ) _UpperCamelCase = jnp.zeros((1, 1, self.cross_attention_dim) , dtype=jnp.floataa ) _UpperCamelCase = (1, 3, self.sample_size * 8, self.sample_size * 8) _UpperCamelCase = jnp.zeros(_A , dtype=jnp.floataa ) _UpperCamelCase , _UpperCamelCase = jax.random.split(_A ) _UpperCamelCase = {'''params''': params_rng, '''dropout''': dropout_rng} return self.init(_A , _A , _A , _A , _A )["params"] def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = self.block_out_channels _UpperCamelCase = block_out_channels[0] * 4 # If `num_attention_heads` is not defined (which is the case for most models) # it will default to `attention_head_dim`. This looks weird upon first reading it and it is. # The reason for this behavior is to correct for incorrectly named variables that were introduced # when this library was created. The incorrect naming was only discovered much later in https://github.com/huggingface/diffusers/issues/2011#issuecomment-1547958131 # Changing `attention_head_dim` to `num_attention_heads` for 40,000+ configurations is too backwards breaking # which is why we correct for the naming here. _UpperCamelCase = self.num_attention_heads or self.attention_head_dim # input _UpperCamelCase = nn.Conv( block_out_channels[0] , kernel_size=(3, 3) , strides=(1, 1) , padding=((1, 1), (1, 1)) , dtype=self.dtype , ) # time _UpperCamelCase = FlaxTimesteps( block_out_channels[0] , flip_sin_to_cos=self.flip_sin_to_cos , freq_shift=self.config.freq_shift ) _UpperCamelCase = FlaxTimestepEmbedding(_A , dtype=self.dtype ) _UpperCamelCase = FlaxControlNetConditioningEmbedding( conditioning_embedding_channels=block_out_channels[0] , block_out_channels=self.conditioning_embedding_out_channels , ) _UpperCamelCase = self.only_cross_attention if isinstance(_A , _A ): _UpperCamelCase = (only_cross_attention,) * len(self.down_block_types ) if isinstance(_A , _A ): _UpperCamelCase = (num_attention_heads,) * len(self.down_block_types ) # down _UpperCamelCase = [] _UpperCamelCase = [] _UpperCamelCase = block_out_channels[0] _UpperCamelCase = nn.Conv( _A , kernel_size=(1, 1) , padding='''VALID''' , kernel_init=nn.initializers.zeros_init() , bias_init=nn.initializers.zeros_init() , dtype=self.dtype , ) controlnet_down_blocks.append(_A ) for i, down_block_type in enumerate(self.down_block_types ): _UpperCamelCase = output_channel _UpperCamelCase = block_out_channels[i] _UpperCamelCase = i == len(_A ) - 1 if down_block_type == "CrossAttnDownBlock2D": _UpperCamelCase = FlaxCrossAttnDownBlockaD( in_channels=_A , out_channels=_A , dropout=self.dropout , num_layers=self.layers_per_block , num_attention_heads=num_attention_heads[i] , add_downsample=not is_final_block , use_linear_projection=self.use_linear_projection , only_cross_attention=only_cross_attention[i] , dtype=self.dtype , ) else: _UpperCamelCase = FlaxDownBlockaD( in_channels=_A , out_channels=_A , dropout=self.dropout , num_layers=self.layers_per_block , add_downsample=not is_final_block , dtype=self.dtype , ) down_blocks.append(_A ) for _ in range(self.layers_per_block ): _UpperCamelCase = nn.Conv( _A , kernel_size=(1, 1) , padding='''VALID''' , kernel_init=nn.initializers.zeros_init() , bias_init=nn.initializers.zeros_init() , dtype=self.dtype , ) controlnet_down_blocks.append(_A ) if not is_final_block: _UpperCamelCase = nn.Conv( _A , kernel_size=(1, 1) , padding='''VALID''' , kernel_init=nn.initializers.zeros_init() , bias_init=nn.initializers.zeros_init() , dtype=self.dtype , ) controlnet_down_blocks.append(_A ) _UpperCamelCase = down_blocks _UpperCamelCase = controlnet_down_blocks # mid _UpperCamelCase = block_out_channels[-1] _UpperCamelCase = FlaxUNetMidBlockaDCrossAttn( in_channels=_A , dropout=self.dropout , num_attention_heads=num_attention_heads[-1] , use_linear_projection=self.use_linear_projection , dtype=self.dtype , ) _UpperCamelCase = nn.Conv( _A , kernel_size=(1, 1) , padding='''VALID''' , kernel_init=nn.initializers.zeros_init() , bias_init=nn.initializers.zeros_init() , dtype=self.dtype , ) def __call__( self : str , _A : Dict , _A : Union[str, Any] , _A : int , _A : Union[str, Any] , _A : float = 1.0 , _A : bool = True , _A : bool = False , ): _UpperCamelCase = self.controlnet_conditioning_channel_order if channel_order == "bgr": _UpperCamelCase = jnp.flip(_A , axis=1 ) # 1. time if not isinstance(_A , jnp.ndarray ): _UpperCamelCase = jnp.array([timesteps] , dtype=jnp.intaa ) elif isinstance(_A , jnp.ndarray ) and len(timesteps.shape ) == 0: _UpperCamelCase = timesteps.astype(dtype=jnp.floataa ) _UpperCamelCase = jnp.expand_dims(_A , 0 ) _UpperCamelCase = self.time_proj(_A ) _UpperCamelCase = self.time_embedding(_A ) # 2. pre-process _UpperCamelCase = jnp.transpose(_A , (0, 2, 3, 1) ) _UpperCamelCase = self.conv_in(_A ) _UpperCamelCase = jnp.transpose(_A , (0, 2, 3, 1) ) _UpperCamelCase = self.controlnet_cond_embedding(_A ) sample += controlnet_cond # 3. down _UpperCamelCase = (sample,) for down_block in self.down_blocks: if isinstance(_A , _A ): _UpperCamelCase , _UpperCamelCase = down_block(_A , _A , _A , deterministic=not train ) else: _UpperCamelCase , _UpperCamelCase = down_block(_A , _A , deterministic=not train ) down_block_res_samples += res_samples # 4. mid _UpperCamelCase = self.mid_block(_A , _A , _A , deterministic=not train ) # 5. contronet blocks _UpperCamelCase = () for down_block_res_sample, controlnet_block in zip(_A , self.controlnet_down_blocks ): _UpperCamelCase = controlnet_block(_A ) controlnet_down_block_res_samples += (down_block_res_sample,) _UpperCamelCase = controlnet_down_block_res_samples _UpperCamelCase = self.controlnet_mid_block(_A ) # 6. scaling _UpperCamelCase = [sample * conditioning_scale for sample in down_block_res_samples] mid_block_res_sample *= conditioning_scale if not return_dict: return (down_block_res_samples, mid_block_res_sample) return FlaxControlNetOutput( down_block_res_samples=_A , mid_block_res_sample=_A )
71
import logging import os import sys from dataclasses import dataclass, field from typing import Optional from seqaseq_trainer import SeqaSeqTrainer from seqaseq_training_args import SeqaSeqTrainingArguments import transformers from transformers import ( AutoConfig, AutoModelForSeqaSeqLM, AutoTokenizer, HfArgumentParser, MBartTokenizer, MBartTokenizerFast, set_seed, ) from transformers.trainer_utils import EvaluationStrategy, is_main_process from transformers.training_args import ParallelMode from utils import ( SeqaSeqDataCollator, SeqaSeqDataset, assert_all_frozen, build_compute_metrics_fn, check_output_dir, freeze_embeds, freeze_params, lmap, save_json, use_task_specific_params, write_txt_file, ) _lowerCAmelCase = logging.getLogger(__name__) @dataclass class lowerCAmelCase_ : UpperCAmelCase = field( metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Pretrained config name or path if not the same as model_name"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"}, ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Whether tp freeze the encoder."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Whether to freeze the embeddings."} ) @dataclass class lowerCAmelCase_ : UpperCAmelCase = field( metadata={"help": "The input data dir. Should contain the .tsv files (or other data files) for the task."} ) UpperCAmelCase = field( default="summarization", metadata={"help": "Task name, summarization (or summarization_{dataset} for pegasus) or translation"}, ) UpperCAmelCase = field( default=1024, metadata={ "help": ( "The maximum total input sequence length after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field( default=128, metadata={ "help": ( "The maximum total sequence length for target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field( default=142, metadata={ "help": ( "The maximum total sequence length for validation target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded. " "This argument is also used to override the ``max_length`` param of ``model.generate``, which is used " "during ``evaluate`` and ``predict``." ) }, ) UpperCAmelCase = field( default=142, metadata={ "help": ( "The maximum total sequence length for test target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field(default=-1, metadata={"help": "# training examples. -1 means use all."} ) UpperCAmelCase = field(default=-1, metadata={"help": "# validation examples. -1 means use all."} ) UpperCAmelCase = field(default=-1, metadata={"help": "# test examples. -1 means use all."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Source language id for translation."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Target language id for translation."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "# num_beams to use for evaluation."} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "If only pad tokens should be ignored. This assumes that `config.pad_token_id` is defined."}, ) def _snake_case ( __snake_case , __snake_case , __snake_case ): logger.info(f"""***** {split} metrics *****""" ) for key in sorted(metrics.keys() ): logger.info(f""" {key} = {metrics[key]}""" ) save_json(__snake_case , os.path.join(__snake_case , f"""{split}_results.json""" ) ) def _snake_case ( ): # See all possible arguments in src/transformers/training_args.py # or by passing the --help flag to this script. # We now keep distinct sets of args, for a cleaner separation of concerns. _UpperCamelCase = HfArgumentParser((ModelArguments, DataTrainingArguments, SeqaSeqTrainingArguments) ) if len(sys.argv ) == 2 and sys.argv[1].endswith('''.json''' ): # If we pass only one argument to the script and it's the path to a json file, # let's parse it to get our arguments. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1] ) ) else: _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = parser.parse_args_into_dataclasses() check_output_dir(__snake_case ) # Setup logging logging.basicConfig( format='''%(asctime)s - %(levelname)s - %(name)s - %(message)s''' , datefmt='''%m/%d/%Y %H:%M:%S''' , level=logging.INFO if training_args.local_rank in [-1, 0] else logging.WARN , ) logger.warning( '''Process rank: %s, device: %s, n_gpu: %s, distributed training: %s, 16-bits training: %s''' , training_args.local_rank , training_args.device , training_args.n_gpu , bool(training_args.parallel_mode == ParallelMode.DISTRIBUTED ) , training_args.fpaa , ) transformers.utils.logging.enable_default_handler() transformers.utils.logging.enable_explicit_format() # Set the verbosity to info of the Transformers logger (on main process only): if is_main_process(training_args.local_rank ): transformers.utils.logging.set_verbosity_info() logger.info('''Training/evaluation parameters %s''' , __snake_case ) # Set seed set_seed(training_args.seed ) # Load pretrained model and tokenizer # # Distributed training: # The .from_pretrained methods guarantee that only one local process can concurrently # download model & vocab. _UpperCamelCase = AutoConfig.from_pretrained( model_args.config_name if model_args.config_name else model_args.model_name_or_path , cache_dir=model_args.cache_dir , ) _UpperCamelCase = ('''encoder_layerdrop''', '''decoder_layerdrop''', '''dropout''', '''attention_dropout''') for p in extra_model_params: if getattr(__snake_case , __snake_case , __snake_case ): assert hasattr(__snake_case , __snake_case ), f"""({config.__class__.__name__}) doesn't have a `{p}` attribute""" setattr(__snake_case , __snake_case , getattr(__snake_case , __snake_case ) ) _UpperCamelCase = AutoTokenizer.from_pretrained( model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path , cache_dir=model_args.cache_dir , ) _UpperCamelCase = AutoModelForSeqaSeqLM.from_pretrained( model_args.model_name_or_path , from_tf='''.ckpt''' in model_args.model_name_or_path , config=__snake_case , cache_dir=model_args.cache_dir , ) # use task specific params use_task_specific_params(__snake_case , data_args.task ) # set num_beams for evaluation if data_args.eval_beams is None: _UpperCamelCase = model.config.num_beams # set decoder_start_token_id for MBart if model.config.decoder_start_token_id is None and isinstance(__snake_case , (MBartTokenizer, MBartTokenizerFast) ): assert ( data_args.tgt_lang is not None and data_args.src_lang is not None ), "mBart requires --tgt_lang and --src_lang" if isinstance(__snake_case , __snake_case ): _UpperCamelCase = tokenizer.lang_code_to_id[data_args.tgt_lang] else: _UpperCamelCase = tokenizer.convert_tokens_to_ids(data_args.tgt_lang ) if model_args.freeze_embeds: freeze_embeds(__snake_case ) if model_args.freeze_encoder: freeze_params(model.get_encoder() ) assert_all_frozen(model.get_encoder() ) _UpperCamelCase = SeqaSeqDataset # Get datasets _UpperCamelCase = ( dataset_class( __snake_case , type_path='''train''' , data_dir=data_args.data_dir , n_obs=data_args.n_train , max_target_length=data_args.max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_train else None ) _UpperCamelCase = ( dataset_class( __snake_case , type_path='''val''' , data_dir=data_args.data_dir , n_obs=data_args.n_val , max_target_length=data_args.val_max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_eval or training_args.evaluation_strategy != EvaluationStrategy.NO else None ) _UpperCamelCase = ( dataset_class( __snake_case , type_path='''test''' , data_dir=data_args.data_dir , n_obs=data_args.n_test , max_target_length=data_args.test_max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_predict else None ) # Initialize our Trainer _UpperCamelCase = ( build_compute_metrics_fn(data_args.task , __snake_case ) if training_args.predict_with_generate else None ) _UpperCamelCase = SeqaSeqTrainer( model=__snake_case , args=__snake_case , data_args=__snake_case , train_dataset=__snake_case , eval_dataset=__snake_case , data_collator=SeqaSeqDataCollator( __snake_case , __snake_case , model.config.decoder_start_token_id , training_args.tpu_num_cores ) , compute_metrics=__snake_case , tokenizer=__snake_case , ) _UpperCamelCase = {} # Training if training_args.do_train: logger.info('''*** Train ***''' ) _UpperCamelCase = trainer.train( model_path=model_args.model_name_or_path if os.path.isdir(model_args.model_name_or_path ) else None ) _UpperCamelCase = train_result.metrics _UpperCamelCase = data_args.n_train trainer.save_model() # this also saves the tokenizer if trainer.is_world_process_zero(): handle_metrics('''train''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) # Need to save the state, since Trainer.save_model saves only the tokenizer with the model trainer.state.save_to_json(os.path.join(training_args.output_dir , '''trainer_state.json''' ) ) # For convenience, we also re-save the tokenizer to the same directory, # so that you can share your model easily on huggingface.co/models =) tokenizer.save_pretrained(training_args.output_dir ) # Evaluation if training_args.do_eval: logger.info('''*** Evaluate ***''' ) _UpperCamelCase = trainer.evaluate(metric_key_prefix='''val''' ) _UpperCamelCase = data_args.n_val _UpperCamelCase = round(metrics['''val_loss'''] , 4 ) if trainer.is_world_process_zero(): handle_metrics('''val''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) if training_args.do_predict: logger.info('''*** Predict ***''' ) _UpperCamelCase = trainer.predict(test_dataset=__snake_case , metric_key_prefix='''test''' ) _UpperCamelCase = test_output.metrics _UpperCamelCase = data_args.n_test if trainer.is_world_process_zero(): _UpperCamelCase = round(metrics['''test_loss'''] , 4 ) handle_metrics('''test''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) if training_args.predict_with_generate: _UpperCamelCase = tokenizer.batch_decode( test_output.predictions , skip_special_tokens=__snake_case , clean_up_tokenization_spaces=__snake_case ) _UpperCamelCase = lmap(str.strip , __snake_case ) write_txt_file(__snake_case , os.path.join(training_args.output_dir , '''test_generations.txt''' ) ) if trainer.is_world_process_zero(): save_json(__snake_case , os.path.join(training_args.output_dir , '''all_results.json''' ) ) return all_metrics def _snake_case ( __snake_case ): # For xla_spawn (TPUs) main() if __name__ == "__main__": main()
71
1
import multiprocessing from typing import TYPE_CHECKING, Optional, Union from .. import Dataset, Features, config from ..formatting import query_table from ..packaged_modules.sql.sql import Sql from ..utils import logging from .abc import AbstractDatasetInputStream if TYPE_CHECKING: import sqlitea import sqlalchemy class lowerCAmelCase_ ( __lowercase ): def __init__( self : Tuple , _A : Union[str, "sqlalchemy.sql.Selectable"] , _A : Union[str, "sqlalchemy.engine.Connection", "sqlalchemy.engine.Engine", "sqlite3.Connection"] , _A : Optional[Features] = None , _A : str = None , _A : bool = False , **_A : Tuple , ): super().__init__(features=_A , cache_dir=_A , keep_in_memory=_A , **_A ) _UpperCamelCase = Sql( cache_dir=_A , features=_A , sql=_A , con=_A , **_A , ) def UpperCamelCase_ ( self : Any ): _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None self.builder.download_and_prepare( download_config=_A , download_mode=_A , verification_mode=_A , base_path=_A , ) # Build dataset for splits _UpperCamelCase = self.builder.as_dataset( split='''train''' , verification_mode=_A , in_memory=self.keep_in_memory ) return dataset class lowerCAmelCase_ : def __init__( self : Union[str, Any] , _A : Dataset , _A : str , _A : Union[str, "sqlalchemy.engine.Connection", "sqlalchemy.engine.Engine", "sqlite3.Connection"] , _A : Optional[int] = None , _A : Optional[int] = None , **_A : Optional[Any] , ): if num_proc is not None and num_proc <= 0: raise ValueError(F"""num_proc {num_proc} must be an integer > 0.""" ) _UpperCamelCase = dataset _UpperCamelCase = name _UpperCamelCase = con _UpperCamelCase = batch_size if batch_size else config.DEFAULT_MAX_BATCH_SIZE _UpperCamelCase = num_proc _UpperCamelCase = to_sql_kwargs def UpperCamelCase_ ( self : Any ): _UpperCamelCase = self.to_sql_kwargs.pop('''sql''' , _A ) _UpperCamelCase = self.to_sql_kwargs.pop('''con''' , _A ) _UpperCamelCase = self.to_sql_kwargs.pop('''index''' , _A ) _UpperCamelCase = self._write(index=_A , **self.to_sql_kwargs ) return written def UpperCamelCase_ ( self : int , _A : Optional[Any] ): _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = args _UpperCamelCase = {**to_sql_kwargs, '''if_exists''': '''append'''} if offset > 0 else to_sql_kwargs _UpperCamelCase = query_table( table=self.dataset.data , key=slice(_A , offset + self.batch_size ) , indices=self.dataset._indices , ) _UpperCamelCase = batch.to_pandas() _UpperCamelCase = df.to_sql(self.name , self.con , index=_A , **_A ) return num_rows or len(_A ) def UpperCamelCase_ ( self : List[str] , _A : Any , **_A : Optional[Any] ): _UpperCamelCase = 0 if self.num_proc is None or self.num_proc == 1: for offset in logging.tqdm( range(0 , len(self.dataset ) , self.batch_size ) , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating SQL from Arrow format''' , ): written += self._batch_sql((offset, index, to_sql_kwargs) ) else: _UpperCamelCase , _UpperCamelCase = len(self.dataset ), self.batch_size with multiprocessing.Pool(self.num_proc ) as pool: for num_rows in logging.tqdm( pool.imap( self._batch_sql , [(offset, index, to_sql_kwargs) for offset in range(0 , _A , _A )] , ) , total=(num_rows // batch_size) + 1 if num_rows % batch_size else num_rows // batch_size , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating SQL from Arrow format''' , ): written += num_rows return written
71
from __future__ import annotations import typing from collections import Counter def _snake_case ( __snake_case ): _UpperCamelCase = Counter() for base in range(1 , max_perimeter + 1 ): for perpendicular in range(__snake_case , max_perimeter + 1 ): _UpperCamelCase = (base * base + perpendicular * perpendicular) ** 0.5 if hypotenuse == int(__snake_case ): _UpperCamelCase = int(base + perpendicular + hypotenuse ) if perimeter > max_perimeter: continue triplets[perimeter] += 1 return triplets def _snake_case ( __snake_case = 1000 ): _UpperCamelCase = pythagorean_triple(__snake_case ) return triplets.most_common(1 )[0][0] if __name__ == "__main__": print(f'Perimeter {solution()} has maximum solutions')
71
1
from typing import TYPE_CHECKING from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available _lowerCAmelCase = { "configuration_jukebox": [ "JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP", "JukeboxConfig", "JukeboxPriorConfig", "JukeboxVQVAEConfig", ], "tokenization_jukebox": ["JukeboxTokenizer"], } try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST", "JukeboxModel", "JukeboxPreTrainedModel", "JukeboxVQVAE", "JukeboxPrior", ] if TYPE_CHECKING: from .configuration_jukebox import ( JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP, JukeboxConfig, JukeboxPriorConfig, JukeboxVQVAEConfig, ) from .tokenization_jukebox import JukeboxTokenizer try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_jukebox import ( JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST, JukeboxModel, JukeboxPreTrainedModel, JukeboxPrior, JukeboxVQVAE, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
import torch from diffusers import DPMSolverSDEScheduler from diffusers.utils import torch_device from diffusers.utils.testing_utils import require_torchsde from .test_schedulers import SchedulerCommonTest @require_torchsde class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = (DPMSolverSDEScheduler,) UpperCAmelCase = 10 def UpperCamelCase_ ( self : Tuple , **_A : Union[str, Any] ): _UpperCamelCase = { '''num_train_timesteps''': 1100, '''beta_start''': 0.0001, '''beta_end''': 0.02, '''beta_schedule''': '''linear''', '''noise_sampler_seed''': 0, } config.update(**_A ) return config def UpperCamelCase_ ( self : List[Any] ): for timesteps in [10, 50, 100, 1000]: self.check_over_configs(num_train_timesteps=_A ) def UpperCamelCase_ ( self : List[Any] ): for beta_start, beta_end in zip([0.0_0001, 0.0001, 0.001] , [0.0002, 0.002, 0.02] ): self.check_over_configs(beta_start=_A , beta_end=_A ) def UpperCamelCase_ ( self : List[str] ): for schedule in ["linear", "scaled_linear"]: self.check_over_configs(beta_schedule=_A ) def UpperCamelCase_ ( self : Union[str, Any] ): for prediction_type in ["epsilon", "v_prediction"]: self.check_over_configs(prediction_type=_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for i, t in enumerate(scheduler.timesteps ): _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 167.47_8210_4492_1875 ) < 1e-2 assert abs(result_mean.item() - 0.2178_7059_6456_5277 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 171.59_3521_1181_6406 ) < 1e-2 assert abs(result_mean.item() - 0.2_2342_9068_9229_9652 ) < 1e-3 else: assert abs(result_sum.item() - 162.52_3834_2285_1562 ) < 1e-2 assert abs(result_mean.item() - 0.211_6195_7085_1326 ) < 1e-3 def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config(prediction_type='''v_prediction''' ) _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for i, t in enumerate(scheduler.timesteps ): _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 124.77_1492_0043_9453 ) < 1e-2 assert abs(result_mean.item() - 0.1_6226_2890_1481_6284 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 128.1_6633_6059_5703 ) < 1e-2 assert abs(result_mean.item() - 0.1_6688_3260_0116_7297 ) < 1e-3 else: assert abs(result_sum.item() - 119.8_4875_4882_8125 ) < 1e-2 assert abs(result_mean.item() - 0.1560_5306_6253_6621 ) < 1e-3 def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps , device=_A ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter.to(_A ) * scheduler.init_noise_sigma for t in scheduler.timesteps: _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 167.46_9573_9746_0938 ) < 1e-2 assert abs(result_mean.item() - 0.2_1805_9346_0798_2635 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 171.59_3536_3769_5312 ) < 1e-2 assert abs(result_mean.item() - 0.2_2342_9083_8241_5771 ) < 1e-3 else: assert abs(result_sum.item() - 162.52_3834_2285_1562 ) < 1e-2 assert abs(result_mean.item() - 0.211_6195_7085_1326 ) < 1e-3 def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A , use_karras_sigmas=_A ) scheduler.set_timesteps(self.num_inference_steps , device=_A ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter.to(_A ) * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for t in scheduler.timesteps: _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 176.66_9741_3574_2188 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 177.63_6535_6445_3125 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2 else: assert abs(result_sum.item() - 170.3_1352_2338_8672 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2
71
1
import os import pytest from datasets import ( get_dataset_config_info, get_dataset_config_names, get_dataset_infos, get_dataset_split_names, inspect_dataset, inspect_metric, ) _lowerCAmelCase = pytest.mark.integration @pytest.mark.parametrize('''path''' , ['''paws''', '''csv'''] ) def _snake_case ( __snake_case , __snake_case ): inspect_dataset(__snake_case , __snake_case ) _UpperCamelCase = path + '''.py''' assert script_name in os.listdir(__snake_case ) assert "__pycache__" not in os.listdir(__snake_case ) @pytest.mark.filterwarnings('''ignore:inspect_metric is deprecated:FutureWarning''' ) @pytest.mark.filterwarnings('''ignore:metric_module_factory is deprecated:FutureWarning''' ) @pytest.mark.parametrize('''path''' , ['''accuracy'''] ) def _snake_case ( __snake_case , __snake_case ): inspect_metric(__snake_case , __snake_case ) _UpperCamelCase = path + '''.py''' assert script_name in os.listdir(__snake_case ) assert "__pycache__" not in os.listdir(__snake_case ) @pytest.mark.parametrize( '''path, config_name, expected_splits''' , [ ('''squad''', '''plain_text''', ['''train''', '''validation''']), ('''dalle-mini/wit''', '''dalle-mini--wit''', ['''train''']), ('''paws''', '''labeled_final''', ['''train''', '''test''', '''validation''']), ] , ) def _snake_case ( __snake_case , __snake_case , __snake_case ): _UpperCamelCase = get_dataset_config_info(__snake_case , config_name=__snake_case ) assert info.config_name == config_name assert list(info.splits.keys() ) == expected_splits @pytest.mark.parametrize( '''path, config_name, expected_exception''' , [ ('''paws''', None, ValueError), ] , ) def _snake_case ( __snake_case , __snake_case , __snake_case ): with pytest.raises(__snake_case ): get_dataset_config_info(__snake_case , config_name=__snake_case ) @pytest.mark.parametrize( '''path, expected''' , [ ('''squad''', '''plain_text'''), ('''acronym_identification''', '''default'''), ('''lhoestq/squad''', '''plain_text'''), ('''lhoestq/test''', '''default'''), ('''lhoestq/demo1''', '''lhoestq--demo1'''), ('''dalle-mini/wit''', '''dalle-mini--wit'''), ] , ) def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = get_dataset_config_names(__snake_case ) assert expected in config_names @pytest.mark.parametrize( '''path, expected_configs, expected_splits_in_first_config''' , [ ('''squad''', ['''plain_text'''], ['''train''', '''validation''']), ('''dalle-mini/wit''', ['''dalle-mini--wit'''], ['''train''']), ('''paws''', ['''labeled_final''', '''labeled_swap''', '''unlabeled_final'''], ['''train''', '''test''', '''validation''']), ] , ) def _snake_case ( __snake_case , __snake_case , __snake_case ): _UpperCamelCase = get_dataset_infos(__snake_case ) assert list(infos.keys() ) == expected_configs _UpperCamelCase = expected_configs[0] assert expected_config in infos _UpperCamelCase = infos[expected_config] assert info.config_name == expected_config assert list(info.splits.keys() ) == expected_splits_in_first_config @pytest.mark.parametrize( '''path, expected_config, expected_splits''' , [ ('''squad''', '''plain_text''', ['''train''', '''validation''']), ('''dalle-mini/wit''', '''dalle-mini--wit''', ['''train''']), ('''paws''', '''labeled_final''', ['''train''', '''test''', '''validation''']), ] , ) def _snake_case ( __snake_case , __snake_case , __snake_case ): _UpperCamelCase = get_dataset_infos(__snake_case ) assert expected_config in infos _UpperCamelCase = infos[expected_config] assert info.config_name == expected_config assert list(info.splits.keys() ) == expected_splits @pytest.mark.parametrize( '''path, config_name, expected_exception''' , [ ('''paws''', None, ValueError), ] , ) def _snake_case ( __snake_case , __snake_case , __snake_case ): with pytest.raises(__snake_case ): get_dataset_split_names(__snake_case , config_name=__snake_case )
71
import unittest from typing import Tuple import torch from diffusers.utils import floats_tensor, randn_tensor, torch_all_close, torch_device from diffusers.utils.testing_utils import require_torch @require_torch class lowerCAmelCase_ : @property def UpperCamelCase_ ( self : Optional[int] ): return self.get_dummy_input() @property def UpperCamelCase_ ( self : Dict ): if self.block_type == "down": return (4, 32, 16, 16) elif self.block_type == "mid": return (4, 32, 32, 32) elif self.block_type == "up": return (4, 32, 64, 64) raise ValueError(F"""'{self.block_type}' is not a supported block_type. Set it to 'up', 'mid', or 'down'.""" ) def UpperCamelCase_ ( self : Union[str, Any] , _A : List[str]=True , _A : Any=False , _A : Union[str, Any]=False , _A : int=False , ): _UpperCamelCase = 4 _UpperCamelCase = 32 _UpperCamelCase = (32, 32) _UpperCamelCase = torch.manual_seed(0 ) _UpperCamelCase = torch.device(_A ) _UpperCamelCase = (batch_size, num_channels) + sizes _UpperCamelCase = randn_tensor(_A , generator=_A , device=_A ) _UpperCamelCase = {'''hidden_states''': hidden_states} if include_temb: _UpperCamelCase = 128 _UpperCamelCase = randn_tensor((batch_size, temb_channels) , generator=_A , device=_A ) if include_res_hidden_states_tuple: _UpperCamelCase = torch.manual_seed(1 ) _UpperCamelCase = (randn_tensor(_A , generator=_A , device=_A ),) if include_encoder_hidden_states: _UpperCamelCase = floats_tensor((batch_size, 32, 32) ).to(_A ) if include_skip_sample: _UpperCamelCase = randn_tensor(((batch_size, 3) + sizes) , generator=_A , device=_A ) return dummy_input def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = { '''in_channels''': 32, '''out_channels''': 32, '''temb_channels''': 128, } if self.block_type == "up": _UpperCamelCase = 32 if self.block_type == "mid": init_dict.pop('''out_channels''' ) _UpperCamelCase = self.dummy_input return init_dict, inputs_dict def UpperCamelCase_ ( self : Tuple , _A : Union[str, Any] ): _UpperCamelCase , _UpperCamelCase = self.prepare_init_args_and_inputs_for_common() _UpperCamelCase = self.block_class(**_A ) unet_block.to(_A ) unet_block.eval() with torch.no_grad(): _UpperCamelCase = unet_block(**_A ) if isinstance(_A , _A ): _UpperCamelCase = output[0] self.assertEqual(output.shape , self.output_shape ) _UpperCamelCase = output[0, -1, -3:, -3:] _UpperCamelCase = torch.tensor(_A ).to(_A ) assert torch_all_close(output_slice.flatten() , _A , atol=5e-3 ) @unittest.skipIf(torch_device == '''mps''' , '''Training is not supported in mps''' ) def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase , _UpperCamelCase = self.prepare_init_args_and_inputs_for_common() _UpperCamelCase = self.block_class(**_A ) model.to(_A ) model.train() _UpperCamelCase = model(**_A ) if isinstance(_A , _A ): _UpperCamelCase = output[0] _UpperCamelCase = torch.device(_A ) _UpperCamelCase = randn_tensor(output.shape , device=_A ) _UpperCamelCase = torch.nn.functional.mse_loss(_A , _A ) loss.backward()
71
1
import json import os import shutil import tempfile import unittest from transformers import BatchEncoding, CanineTokenizer from transformers.testing_utils import require_tokenizers, require_torch from transformers.tokenization_utils import AddedToken from transformers.utils import cached_property from ...test_tokenization_common import TokenizerTesterMixin class lowerCAmelCase_ ( __lowercase, unittest.TestCase ): UpperCAmelCase = CanineTokenizer UpperCAmelCase = False def UpperCamelCase_ ( self : List[str] ): super().setUp() _UpperCamelCase = CanineTokenizer() tokenizer.save_pretrained(self.tmpdirname ) @cached_property def UpperCamelCase_ ( self : int ): return CanineTokenizer.from_pretrained('''google/canine-s''' ) def UpperCamelCase_ ( self : int , **_A : Dict ): _UpperCamelCase = self.tokenizer_class.from_pretrained(self.tmpdirname , **_A ) _UpperCamelCase = 1024 return tokenizer @require_torch def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = self.canine_tokenizer _UpperCamelCase = ['''Life is like a box of chocolates.''', '''You never know what you\'re gonna get.'''] # fmt: off _UpperCamelCase = [5_7344, 76, 105, 102, 101, 32, 105, 115, 32, 108, 105, 107, 101, 32, 97, 32, 98, 111, 120, 32, 111, 102, 32, 99, 104, 111, 99, 111, 108, 97, 116, 101, 115, 46, 5_7345, 0, 0, 0, 0] # fmt: on _UpperCamelCase = tokenizer(_A , padding=_A , return_tensors='''pt''' ) self.assertIsInstance(_A , _A ) _UpperCamelCase = list(batch.input_ids.numpy()[0] ) self.assertListEqual(_A , _A ) self.assertEqual((2, 39) , batch.input_ids.shape ) self.assertEqual((2, 39) , batch.attention_mask.shape ) @require_torch def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.canine_tokenizer _UpperCamelCase = ['''Once there was a man.''', '''He wrote a test in HuggingFace Tranformers.'''] _UpperCamelCase = tokenizer(_A , padding=_A , return_tensors='''pt''' ) # check if input_ids, attention_mask and token_type_ids are returned self.assertIn('''input_ids''' , _A ) self.assertIn('''attention_mask''' , _A ) self.assertIn('''token_type_ids''' , _A ) @require_torch def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.canine_tokenizer _UpperCamelCase = [ '''What\'s the weater?''', '''It\'s about 25 degrees.''', ] _UpperCamelCase = tokenizer( text_target=_A , max_length=32 , padding='''max_length''' , truncation=_A , return_tensors='''pt''' ) self.assertEqual(32 , targets['''input_ids'''].shape[1] ) def UpperCamelCase_ ( self : Any ): # safety check on max_len default value so we are sure the test works _UpperCamelCase = self.get_tokenizers() for tokenizer in tokenizers: with self.subTest(F"""{tokenizer.__class__.__name__}""" ): self.assertNotEqual(tokenizer.model_max_length , 42 ) # Now let's start the test _UpperCamelCase = self.get_tokenizers() for tokenizer in tokenizers: with self.subTest(F"""{tokenizer.__class__.__name__}""" ): # Isolate this from the other tests because we save additional tokens/etc _UpperCamelCase = tempfile.mkdtemp() _UpperCamelCase = ''' He is very happy, UNwant\u00E9d,running''' _UpperCamelCase = tokenizer.encode(_A , add_special_tokens=_A ) tokenizer.save_pretrained(_A ) _UpperCamelCase = tokenizer.__class__.from_pretrained(_A ) _UpperCamelCase = after_tokenizer.encode(_A , add_special_tokens=_A ) self.assertListEqual(_A , _A ) shutil.rmtree(_A ) _UpperCamelCase = self.get_tokenizers(model_max_length=42 ) for tokenizer in tokenizers: with self.subTest(F"""{tokenizer.__class__.__name__}""" ): # Isolate this from the other tests because we save additional tokens/etc _UpperCamelCase = tempfile.mkdtemp() _UpperCamelCase = ''' He is very happy, UNwant\u00E9d,running''' _UpperCamelCase = tokenizer.additional_special_tokens # We can add a new special token for Canine as follows: _UpperCamelCase = chr(0xe_007 ) additional_special_tokens.append(_A ) tokenizer.add_special_tokens({'''additional_special_tokens''': additional_special_tokens} ) _UpperCamelCase = tokenizer.encode(_A , add_special_tokens=_A ) tokenizer.save_pretrained(_A ) _UpperCamelCase = tokenizer.__class__.from_pretrained(_A ) _UpperCamelCase = after_tokenizer.encode(_A , add_special_tokens=_A ) self.assertListEqual(_A , _A ) self.assertIn(_A , after_tokenizer.additional_special_tokens ) self.assertEqual(after_tokenizer.model_max_length , 42 ) _UpperCamelCase = tokenizer.__class__.from_pretrained(_A , model_max_length=43 ) self.assertEqual(tokenizer.model_max_length , 43 ) shutil.rmtree(_A ) def UpperCamelCase_ ( self : str ): _UpperCamelCase = self.get_tokenizers(do_lower_case=_A ) for tokenizer in tokenizers: with self.subTest(F"""{tokenizer.__class__.__name__}""" ): _UpperCamelCase , _UpperCamelCase = self.get_clean_sequence(_A ) # a special token for Canine can be defined as follows: _UpperCamelCase = 0xe_005 _UpperCamelCase = chr(_A ) tokenizer.add_special_tokens({'''cls_token''': special_token} ) _UpperCamelCase = tokenizer.encode(_A , add_special_tokens=_A ) self.assertEqual(len(_A ) , 1 ) _UpperCamelCase = tokenizer.decode(ids + encoded_special_token , clean_up_tokenization_spaces=_A ) _UpperCamelCase = tokenizer.encode(_A , add_special_tokens=_A ) _UpperCamelCase = tokenizer.encode(_A , add_special_tokens=_A ) _UpperCamelCase = tokenizer.encode(_A , add_special_tokens=_A ) self.assertEqual(_A , input_encoded + special_token_id ) _UpperCamelCase = tokenizer.decode(_A , skip_special_tokens=_A ) self.assertTrue(special_token not in decoded ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.get_tokenizers(do_lower_case=_A ) for tokenizer in tokenizers: with self.subTest(F"""{tokenizer.__class__.__name__}""" ): _UpperCamelCase = chr(0xe_005 ) _UpperCamelCase = chr(0xe_006 ) # `add_tokens` method stores special tokens only in `tokenizer.unique_no_split_tokens`. (in tokenization_utils.py) tokenizer.add_tokens([SPECIAL_TOKEN_1] , special_tokens=_A ) # `add_special_tokens` method stores special tokens in `tokenizer.additional_special_tokens`, # which also occur in `tokenizer.all_special_tokens`. (in tokenization_utils_base.py) tokenizer.add_special_tokens({'''additional_special_tokens''': [SPECIAL_TOKEN_2]} ) _UpperCamelCase = tokenizer.tokenize(_A ) _UpperCamelCase = tokenizer.tokenize(_A ) self.assertEqual(len(_A ) , 1 ) self.assertEqual(len(_A ) , 1 ) self.assertEqual(token_a[0] , _A ) self.assertEqual(token_a[0] , _A ) @require_tokenizers def UpperCamelCase_ ( self : str ): _UpperCamelCase = self.get_tokenizers(do_lower_case=_A ) for tokenizer in tokenizers: with self.subTest(F"""{tokenizer.__class__.__name__}""" ): # a special token for Canine can be defined as follows: _UpperCamelCase = 0xe_006 _UpperCamelCase = chr(_A ) _UpperCamelCase = AddedToken(_A , lstrip=_A ) tokenizer.add_special_tokens({'''additional_special_tokens''': [new_token]} ) with tempfile.TemporaryDirectory() as tmp_dir_name: tokenizer.save_pretrained(_A ) tokenizer.from_pretrained(_A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = [] if self.test_slow_tokenizer: tokenizer_list.append((self.tokenizer_class, self.get_tokenizer()) ) if self.test_rust_tokenizer: tokenizer_list.append((self.rust_tokenizer_class, self.get_rust_tokenizer()) ) for tokenizer_class, tokenizer_utils in tokenizer_list: with tempfile.TemporaryDirectory() as tmp_dir: tokenizer_utils.save_pretrained(_A ) with open(os.path.join(_A , '''special_tokens_map.json''' ) , encoding='''utf-8''' ) as json_file: _UpperCamelCase = json.load(_A ) with open(os.path.join(_A , '''tokenizer_config.json''' ) , encoding='''utf-8''' ) as json_file: _UpperCamelCase = json.load(_A ) # a special token for Canine can be defined as follows: _UpperCamelCase = 0xe_006 _UpperCamelCase = chr(_A ) _UpperCamelCase = [new_token_a] _UpperCamelCase = [new_token_a] with open(os.path.join(_A , '''special_tokens_map.json''' ) , '''w''' , encoding='''utf-8''' ) as outfile: json.dump(_A , _A ) with open(os.path.join(_A , '''tokenizer_config.json''' ) , '''w''' , encoding='''utf-8''' ) as outfile: json.dump(_A , _A ) # the following checks allow us to verify that our test works as expected, i.e. that the tokenizer takes # into account the new value of additional_special_tokens given in the "tokenizer_config.json" and # "special_tokens_map.json" files _UpperCamelCase = tokenizer_class.from_pretrained(_A , extra_ids=0 ) self.assertIn(_A , tokenizer_without_change_in_init.additional_special_tokens ) # self.assertIn("an_additional_special_token",tokenizer_without_change_in_init.get_vocab()) # ByT5Tokenization no vocab self.assertEqual( [new_token_a] , tokenizer_without_change_in_init.convert_ids_to_tokens( tokenizer_without_change_in_init.convert_tokens_to_ids([new_token_a] ) ) , ) _UpperCamelCase = 0xe_007 _UpperCamelCase = chr(_A ) # Now we test that we can change the value of additional_special_tokens in the from_pretrained _UpperCamelCase = [AddedToken(_A , lstrip=_A )] _UpperCamelCase = tokenizer_class.from_pretrained( _A , additional_special_tokens=_A , extra_ids=0 ) self.assertIn(_A , tokenizer.additional_special_tokens ) # self.assertIn(new_token_2,tokenizer.get_vocab()) # ByT5Tokenization no vocab self.assertEqual( [new_token_a] , tokenizer.convert_ids_to_tokens(tokenizer.convert_tokens_to_ids([new_token_a] ) ) ) @require_tokenizers def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = self.get_tokenizers(do_lower_case=_A ) for tokenizer in tokenizers: with self.subTest(F"""{tokenizer.__class__.__name__}""" ): _UpperCamelCase = '''hello world''' if self.space_between_special_tokens: _UpperCamelCase = '''[CLS] hello world [SEP]''' else: _UpperCamelCase = input _UpperCamelCase = tokenizer.encode(_A , add_special_tokens=_A ) _UpperCamelCase = tokenizer.decode(_A , spaces_between_special_tokens=self.space_between_special_tokens ) self.assertIn(_A , [output, output.lower()] ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.get_tokenizers() for tokenizer in tokenizers: with self.subTest(F"""{tokenizer.__class__.__name__}""" ): _UpperCamelCase = [ '''bos_token''', '''eos_token''', '''unk_token''', '''sep_token''', '''pad_token''', '''cls_token''', '''mask_token''', ] _UpperCamelCase = '''a''' _UpperCamelCase = ord(_A ) for attr in attributes_list: setattr(_A , attr + '''_id''' , _A ) self.assertEqual(getattr(_A , _A ) , _A ) self.assertEqual(getattr(_A , attr + '''_id''' ) , _A ) setattr(_A , attr + '''_id''' , _A ) self.assertEqual(getattr(_A , _A ) , _A ) self.assertEqual(getattr(_A , attr + '''_id''' ) , _A ) setattr(_A , '''additional_special_tokens_ids''' , [] ) self.assertListEqual(getattr(_A , '''additional_special_tokens''' ) , [] ) self.assertListEqual(getattr(_A , '''additional_special_tokens_ids''' ) , [] ) _UpperCamelCase = 0xe_006 _UpperCamelCase = chr(_A ) setattr(_A , '''additional_special_tokens_ids''' , [additional_special_token_id] ) self.assertListEqual(getattr(_A , '''additional_special_tokens''' ) , [additional_special_token] ) self.assertListEqual(getattr(_A , '''additional_special_tokens_ids''' ) , [additional_special_token_id] ) def UpperCamelCase_ ( self : str ): pass def UpperCamelCase_ ( self : str ): pass def UpperCamelCase_ ( self : Dict ): pass def UpperCamelCase_ ( self : Optional[int] ): pass def UpperCamelCase_ ( self : Union[str, Any] ): pass def UpperCamelCase_ ( self : Dict ): pass def UpperCamelCase_ ( self : str ): pass def UpperCamelCase_ ( self : Optional[Any] ): pass
71
def _snake_case ( __snake_case ): if not isinstance(__snake_case , __snake_case ): raise TypeError('''Input value must be an \'int\' type''' ) _UpperCamelCase = 0 while number: position += 1 number >>= 1 return position if __name__ == "__main__": import doctest doctest.testmod()
71
1
_lowerCAmelCase = { "joule": 1.0, "kilojoule": 1_000, "megajoule": 1_000_000, "gigajoule": 1_000_000_000, "wattsecond": 1.0, "watthour": 3_600, "kilowatthour": 3_600_000, "newtonmeter": 1.0, "calorie_nutr": 4_186.8, "kilocalorie_nutr": 4_186_800.00, "electronvolt": 1.602176634E-19, "britishthermalunit_it": 1_055.05_585, "footpound": 1.355818, } def _snake_case ( __snake_case , __snake_case , __snake_case ): if to_type not in ENERGY_CONVERSION or from_type not in ENERGY_CONVERSION: _UpperCamelCase = ( f"""Incorrect 'from_type' or 'to_type' value: {from_type!r}, {to_type!r}\n""" f"""Valid values are: {", ".join(__snake_case )}""" ) raise ValueError(__snake_case ) return value * ENERGY_CONVERSION[from_type] / ENERGY_CONVERSION[to_type] if __name__ == "__main__": import doctest doctest.testmod()
71
import argparse import json import os import fairseq import torch from torch import nn from transformers import ( SpeechaTextaConfig, SpeechaTextaForCausalLM, SpeechaTextaTokenizer, SpeechEncoderDecoderConfig, SpeechEncoderDecoderModel, WavaVecaConfig, WavaVecaFeatureExtractor, WavaVecaModel, logging, ) logging.set_verbosity_info() _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "post_extract_proj": "feature_projection.projection", "encoder.pos_conv.0": "encoder.pos_conv_embed.conv", "self_attn.k_proj": "encoder.layers.*.attention.k_proj", "self_attn.v_proj": "encoder.layers.*.attention.v_proj", "self_attn.q_proj": "encoder.layers.*.attention.q_proj", "self_attn.out_proj": "encoder.layers.*.attention.out_proj", "self_attn_layer_norm": "encoder.layers.*.layer_norm", "fc1": "encoder.layers.*.feed_forward.intermediate_dense", "fc2": "encoder.layers.*.feed_forward.output_dense", "final_layer_norm": "encoder.layers.*.final_layer_norm", "encoder.layer_norm": "encoder.layer_norm", "w2v_model.layer_norm": "feature_projection.layer_norm", "quantizer.weight_proj": "quantizer.weight_proj", "quantizer.vars": "quantizer.codevectors", "project_q": "project_q", "final_proj": "project_hid", "w2v_encoder.proj": "lm_head", "mask_emb": "masked_spec_embed", } _lowerCAmelCase = [ "lm_head", "quantizer.weight_proj", "quantizer.codevectors", "project_q", "project_hid", ] def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): for attribute in key.split('''.''' ): _UpperCamelCase = getattr(__snake_case , __snake_case ) if weight_type is not None: _UpperCamelCase = getattr(__snake_case , __snake_case ).shape else: _UpperCamelCase = hf_pointer.shape assert hf_shape == value.shape, ( f"""Shape of hf {key + "." + weight_type if weight_type is not None else ""} is {hf_shape}, but should be""" f""" {value.shape} for {full_name}""" ) if weight_type == "weight": _UpperCamelCase = value elif weight_type == "weight_g": _UpperCamelCase = value elif weight_type == "weight_v": _UpperCamelCase = value elif weight_type == "bias": _UpperCamelCase = value else: _UpperCamelCase = value logger.info(f"""{key + "." + weight_type if weight_type is not None else ""} was initialized from {full_name}.""" ) def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = [] _UpperCamelCase = fairseq_model.state_dict() _UpperCamelCase = hf_model.feature_extractor # if encoder has different dim to decoder -> use proj_weight _UpperCamelCase = None for name, value in fairseq_dict.items(): _UpperCamelCase = False if "conv_layers" in name: load_conv_layer( __snake_case , __snake_case , __snake_case , __snake_case , hf_model.config.feat_extract_norm == '''group''' , ) _UpperCamelCase = True elif name.split('''.''' )[0] == "proj": _UpperCamelCase = fairseq_model.proj _UpperCamelCase = True else: for key, mapped_key in MAPPING.items(): if key in name or key.split('''w2v_model.''' )[-1] == name.split('''.''' )[0]: _UpperCamelCase = True if "*" in mapped_key: _UpperCamelCase = name.split(__snake_case )[0].split('''.''' )[-2] _UpperCamelCase = mapped_key.replace('''*''' , __snake_case ) if "weight_g" in name: _UpperCamelCase = '''weight_g''' elif "weight_v" in name: _UpperCamelCase = '''weight_v''' elif "bias" in name: _UpperCamelCase = '''bias''' elif "weight" in name: _UpperCamelCase = '''weight''' else: _UpperCamelCase = None set_recursively(__snake_case , __snake_case , __snake_case , __snake_case , __snake_case ) continue if not is_used: unused_weights.append(__snake_case ) logger.warning(f"""Unused weights: {unused_weights}""" ) return proj_weight def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = full_name.split('''conv_layers.''' )[-1] _UpperCamelCase = name.split('''.''' ) _UpperCamelCase = int(items[0] ) _UpperCamelCase = int(items[1] ) if type_id == 0: if "bias" in name: assert value.shape == feature_extractor.conv_layers[layer_id].conv.bias.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor.conv_layers[layer_id].conv.bias.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract conv layer {layer_id} was initialized from {full_name}.""" ) elif "weight" in name: assert value.shape == feature_extractor.conv_layers[layer_id].conv.weight.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor.conv_layers[layer_id].conv.weight.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract conv layer {layer_id} was initialized from {full_name}.""" ) elif (type_id == 2 and not use_group_norm) or (type_id == 2 and layer_id == 0 and use_group_norm): if "bias" in name: assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.bias.data.shape, ( f"""{full_name} has size {value.shape}, but {feature_extractor[layer_id].layer_norm.bias.data.shape} was""" " found." ) _UpperCamelCase = value logger.info(f"""Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.""" ) elif "weight" in name: assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.weight.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor[layer_id].layer_norm.weight.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.""" ) else: unused_weights.append(__snake_case ) def _snake_case ( __snake_case ): _UpperCamelCase , _UpperCamelCase = emb.weight.shape _UpperCamelCase = nn.Linear(__snake_case , __snake_case , bias=__snake_case ) _UpperCamelCase = emb.weight.data return lin_layer def _snake_case ( __snake_case ): with open(__snake_case , '''r''' , encoding='''utf-8''' ) as f: _UpperCamelCase = f.readlines() _UpperCamelCase = [line.split(''' ''' )[0] for line in lines] _UpperCamelCase = len(__snake_case ) _UpperCamelCase = { '''<s>''': 0, '''<pad>''': 1, '''</s>''': 2, '''<unk>''': 3, } vocab_dict.update(dict(zip(__snake_case , range(4 , num_words + 4 ) ) ) ) return vocab_dict @torch.no_grad() def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , ): _UpperCamelCase = WavaVecaConfig.from_pretrained(__snake_case ) _UpperCamelCase = SpeechaTextaConfig.from_pretrained( __snake_case , vocab_size=__snake_case , decoder_layers=__snake_case , do_stable_layer_norm=__snake_case ) _UpperCamelCase = WavaVecaFeatureExtractor( feature_size=1 , sampling_rate=16000 , padding_value=0 , do_normalize=__snake_case , return_attention_mask=__snake_case , ) _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = fairseq.checkpoint_utils.load_model_ensemble_and_task( [checkpoint_path] , arg_overrides={'''data''': '''/'''.join(dict_path.split('''/''' )[:-1] )} ) _UpperCamelCase = model[0].eval() # set weights for wav2vec2 encoder _UpperCamelCase = WavaVecaModel(__snake_case ) _UpperCamelCase = recursively_load_weights_wavaveca(model.encoder , __snake_case ) _UpperCamelCase = SpeechaTextaForCausalLM(__snake_case ) _UpperCamelCase , _UpperCamelCase = hf_decoder.model.decoder.load_state_dict(model.decoder.state_dict() , strict=__snake_case ) # set output linear layer unexpected_keys.remove('''embed_out''' ) _UpperCamelCase = nn.Parameter(model.decoder.embed_out.detach() ) # layer norm is init to identity matrix so leaving it is fine logger.warning(f"""The following keys are missing when loading the decoder weights: {missing_keys}""" ) logger.warning(f"""The following keys are unexpected when loading the decoder weights: {unexpected_keys}""" ) _UpperCamelCase = SpeechEncoderDecoderModel(encoder=__snake_case , decoder=__snake_case ) _UpperCamelCase = False # add projection layer _UpperCamelCase = nn.Parameter(projection_layer.weight ) _UpperCamelCase = nn.Parameter(projection_layer.bias ) _UpperCamelCase = create_vocab_dict(__snake_case ) with open(os.path.join(__snake_case , '''vocab.json''' ) , '''w''' ) as fp: json.dump(__snake_case , __snake_case ) _UpperCamelCase = SpeechaTextaTokenizer(os.path.join(__snake_case , '''vocab.json''' ) ) tokenizer.save_pretrained(__snake_case ) _UpperCamelCase = hf_wavavec.config.to_dict() _UpperCamelCase = tokenizer.pad_token_id _UpperCamelCase = tokenizer.bos_token_id _UpperCamelCase = tokenizer.eos_token_id _UpperCamelCase = '''speech_to_text_2''' _UpperCamelCase = '''wav2vec2''' _UpperCamelCase = SpeechEncoderDecoderConfig.from_dict(__snake_case ) hf_wavavec.save_pretrained(__snake_case ) feature_extractor.save_pretrained(__snake_case ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() parser.add_argument("--pytorch_dump_folder_path", default=None, type=str, help="Path to the output PyTorch model.") parser.add_argument("--checkpoint_path", default=None, type=str, help="Path to fairseq checkpoint") parser.add_argument("--dict_path", default=None, type=str, help="Path to dict of fine-tuned model") parser.add_argument( "--encoder_config_path", default="facebook/wav2vec2-large-lv60", type=str, help="Path to hf encoder wav2vec2 checkpoint config", ) parser.add_argument( "--decoder_config_path", default="facebook/s2t-small-mustc-en-fr-st", type=str, help="Path to hf decoder s2t checkpoint config", ) parser.add_argument("--vocab_size", default=10_224, type=int, help="Vocab size of decoder") parser.add_argument("--num_decoder_layers", default=7, type=int, help="Number of decoder layers") _lowerCAmelCase = parser.parse_args() convert_wavaveca_checkpoint( args.checkpoint_path, args.pytorch_dump_folder_path, args.dict_path, encoder_config_path=args.encoder_config_path, decoder_config_path=args.decoder_config_path, vocab_size=args.vocab_size, num_decoder_layers=args.num_decoder_layers, )
71
1
from typing import Dict, List, Optional, Union import numpy as np from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict from ...image_transforms import ( center_crop, get_resize_output_image_size, normalize, rescale, resize, to_channel_dimension_format, ) from ...image_utils import ( IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD, ChannelDimension, ImageInput, PILImageResampling, make_list_of_images, to_numpy_array, valid_images, ) from ...utils import TensorType, is_vision_available, logging if is_vision_available(): import PIL _lowerCAmelCase = logging.get_logger(__name__) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = ["pixel_values"] def __init__( self : int , _A : bool = True , _A : Dict[str, int] = None , _A : int = 0.9 , _A : PILImageResampling = PILImageResampling.BICUBIC , _A : bool = True , _A : Dict[str, int] = None , _A : Union[int, float] = 1 / 255 , _A : bool = True , _A : bool = True , _A : Optional[Union[float, List[float]]] = None , _A : Optional[Union[float, List[float]]] = None , **_A : Any , ): super().__init__(**_A ) _UpperCamelCase = size if size is not None else {'''shortest_edge''': 224} _UpperCamelCase = get_size_dict(_A , default_to_square=_A ) _UpperCamelCase = crop_size if crop_size is not None else {'''height''': 224, '''width''': 224} _UpperCamelCase = get_size_dict(_A , param_name='''crop_size''' ) _UpperCamelCase = do_resize _UpperCamelCase = size _UpperCamelCase = crop_pct _UpperCamelCase = resample _UpperCamelCase = do_center_crop _UpperCamelCase = crop_size _UpperCamelCase = do_rescale _UpperCamelCase = rescale_factor _UpperCamelCase = do_normalize _UpperCamelCase = image_mean if image_mean is not None else IMAGENET_DEFAULT_MEAN _UpperCamelCase = image_std if image_std is not None else IMAGENET_DEFAULT_STD def UpperCamelCase_ ( self : int , _A : np.ndarray , _A : Dict[str, int] , _A : Optional[float] = None , _A : PILImageResampling = PILImageResampling.BICUBIC , _A : Optional[Union[str, ChannelDimension]] = None , **_A : Union[str, Any] , ): _UpperCamelCase = get_size_dict(_A , default_to_square=_A ) if "shortest_edge" not in size and ("height" not in size or "width" not in size): raise ValueError(F"""size must contain 'height' and 'width' or 'shortest_edge' as keys. Got {size.keys()}""" ) if crop_pct is not None: if "shortest_edge" in size: _UpperCamelCase = int(size['''shortest_edge'''] / crop_pct ) elif "height" in size and "width" in size: if size["height"] == size["width"]: _UpperCamelCase = int(size['''height'''] / crop_pct ) else: _UpperCamelCase = (int(size['''height'''] / crop_pct ), int(size['''width'''] / crop_pct )) else: raise ValueError('''Invalid size for resize: {}'''.format(_A ) ) _UpperCamelCase = get_resize_output_image_size(_A , size=_A , default_to_square=_A ) else: if "shortest_edge" in size: _UpperCamelCase = get_resize_output_image_size(_A , size=size['''shortest_edge'''] , default_to_square=_A ) elif "height" in size and "width" in size: _UpperCamelCase = (size['''height'''], size['''width''']) else: raise ValueError('''Invalid size for resize: {}'''.format(_A ) ) return resize(_A , size=_A , resample=_A , data_format=_A , **_A ) def UpperCamelCase_ ( self : List[Any] , _A : np.ndarray , _A : Dict[str, int] , _A : Optional[Union[str, ChannelDimension]] = None , **_A : Union[str, Any] , ): _UpperCamelCase = get_size_dict(_A ) if "height" not in size or "width" not in size: raise ValueError(F"""size must contain 'height' and 'width' as keys. Got {size.keys()}""" ) return center_crop(_A , size=(size['''height'''], size['''width''']) , data_format=_A , **_A ) def UpperCamelCase_ ( self : Dict , _A : np.ndarray , _A : Union[int, float] , _A : Optional[Union[str, ChannelDimension]] = None , **_A : Tuple , ): return rescale(_A , scale=_A , data_format=_A , **_A ) def UpperCamelCase_ ( self : Any , _A : np.ndarray , _A : Union[float, List[float]] , _A : Union[float, List[float]] , _A : Optional[Union[str, ChannelDimension]] = None , **_A : Tuple , ): return normalize(_A , mean=_A , std=_A , data_format=_A , **_A ) def UpperCamelCase_ ( self : List[Any] , _A : ImageInput , _A : bool = None , _A : Dict[str, int] = None , _A : int = None , _A : PILImageResampling = None , _A : bool = None , _A : Dict[str, int] = None , _A : bool = None , _A : float = None , _A : bool = None , _A : Optional[Union[float, List[float]]] = None , _A : Optional[Union[float, List[float]]] = None , _A : Optional[Union[str, TensorType]] = None , _A : ChannelDimension = ChannelDimension.FIRST , **_A : Dict , ): _UpperCamelCase = do_resize if do_resize is not None else self.do_resize _UpperCamelCase = crop_pct if crop_pct is not None else self.crop_pct _UpperCamelCase = resample if resample is not None else self.resample _UpperCamelCase = do_center_crop if do_center_crop is not None else self.do_center_crop _UpperCamelCase = do_rescale if do_rescale is not None else self.do_rescale _UpperCamelCase = rescale_factor if rescale_factor is not None else self.rescale_factor _UpperCamelCase = do_normalize if do_normalize is not None else self.do_normalize _UpperCamelCase = image_mean if image_mean is not None else self.image_mean _UpperCamelCase = image_std if image_std is not None else self.image_std _UpperCamelCase = size if size is not None else self.size _UpperCamelCase = get_size_dict(_A , default_to_square=_A ) _UpperCamelCase = crop_size if crop_size is not None else self.crop_size _UpperCamelCase = get_size_dict(_A , param_name='''crop_size''' ) _UpperCamelCase = make_list_of_images(_A ) if not valid_images(_A ): raise ValueError( '''Invalid image type. Must be of type PIL.Image.Image, numpy.ndarray, ''' '''torch.Tensor, tf.Tensor or jax.ndarray.''' ) if do_resize and size is None or resample is None: raise ValueError('''Size and resample must be specified if do_resize is True.''' ) if do_center_crop and crop_pct is None: raise ValueError('''Crop_pct must be specified if do_center_crop is True.''' ) if do_rescale and rescale_factor is None: raise ValueError('''Rescale factor must be specified if do_rescale is True.''' ) if do_normalize and (image_mean is None or image_std is None): raise ValueError('''Image mean and std must be specified if do_normalize is True.''' ) # All transformations expect numpy arrays. _UpperCamelCase = [to_numpy_array(_A ) for image in images] if do_resize: _UpperCamelCase = [self.resize(image=_A , size=_A , crop_pct=_A , resample=_A ) for image in images] if do_center_crop: _UpperCamelCase = [self.center_crop(image=_A , size=_A ) for image in images] if do_rescale: _UpperCamelCase = [self.rescale(image=_A , scale=_A ) for image in images] if do_normalize: _UpperCamelCase = [self.normalize(image=_A , mean=_A , std=_A ) for image in images] _UpperCamelCase = [to_channel_dimension_format(_A , _A ) for image in images] _UpperCamelCase = {'''pixel_values''': images} return BatchFeature(data=_A , tensor_type=_A )
71
from __future__ import annotations import unittest from transformers import DebertaVaConfig, is_tf_available from transformers.testing_utils import require_tf, slow from ...test_configuration_common import ConfigTester from ...test_modeling_tf_common import TFModelTesterMixin, ids_tensor, random_attention_mask from ...test_pipeline_mixin import PipelineTesterMixin if is_tf_available(): import tensorflow as tf from transformers import ( TFDebertaVaForMaskedLM, TFDebertaVaForQuestionAnswering, TFDebertaVaForSequenceClassification, TFDebertaVaForTokenClassification, TFDebertaVaModel, ) class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : Optional[Any] , _A : List[str]=13 , _A : Union[str, Any]=7 , _A : int=True , _A : Optional[int]=True , _A : Optional[int]=True , _A : Union[str, Any]=True , _A : Optional[int]=99 , _A : Union[str, Any]=32 , _A : Dict=2 , _A : List[Any]=4 , _A : Optional[Any]=37 , _A : int="gelu" , _A : Optional[int]=0.1 , _A : str=0.1 , _A : List[str]=512 , _A : Optional[Any]=16 , _A : Optional[Any]=2 , _A : Optional[int]=0.02 , _A : str=False , _A : int=True , _A : Any="None" , _A : Dict=3 , _A : List[Any]=4 , _A : Optional[Any]=None , ): _UpperCamelCase = parent _UpperCamelCase = batch_size _UpperCamelCase = seq_length _UpperCamelCase = is_training _UpperCamelCase = use_input_mask _UpperCamelCase = use_token_type_ids _UpperCamelCase = use_labels _UpperCamelCase = vocab_size _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = hidden_dropout_prob _UpperCamelCase = attention_probs_dropout_prob _UpperCamelCase = max_position_embeddings _UpperCamelCase = type_vocab_size _UpperCamelCase = type_sequence_label_size _UpperCamelCase = initializer_range _UpperCamelCase = num_labels _UpperCamelCase = num_choices _UpperCamelCase = relative_attention _UpperCamelCase = position_biased_input _UpperCamelCase = pos_att_type _UpperCamelCase = scope def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size ) _UpperCamelCase = None if self.use_input_mask: _UpperCamelCase = random_attention_mask([self.batch_size, self.seq_length] ) _UpperCamelCase = None if self.use_token_type_ids: _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.type_vocab_size ) _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None if self.use_labels: _UpperCamelCase = ids_tensor([self.batch_size] , self.type_sequence_label_size ) _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.num_labels ) _UpperCamelCase = DebertaVaConfig( vocab_size=self.vocab_size , hidden_size=self.hidden_size , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , relative_attention=self.relative_attention , position_biased_input=self.position_biased_input , initializer_range=self.initializer_range , return_dict=_A , ) return config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels def UpperCamelCase_ ( self : Dict , _A : Tuple , _A : Tuple , _A : Union[str, Any] , _A : List[str] , _A : Optional[int] , _A : int , _A : Optional[Any] ): _UpperCamelCase = TFDebertaVaModel(config=_A ) _UpperCamelCase = {'''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids} _UpperCamelCase = [input_ids, input_mask] _UpperCamelCase = model(_A ) _UpperCamelCase = model(_A ) self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) ) def UpperCamelCase_ ( self : Dict , _A : Optional[int] , _A : Any , _A : Dict , _A : Union[str, Any] , _A : Union[str, Any] , _A : List[Any] , _A : List[str] ): _UpperCamelCase = TFDebertaVaForMaskedLM(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) ) def UpperCamelCase_ ( self : Dict , _A : Dict , _A : List[str] , _A : List[Any] , _A : List[Any] , _A : Optional[Any] , _A : Tuple , _A : int ): _UpperCamelCase = self.num_labels _UpperCamelCase = TFDebertaVaForSequenceClassification(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) ) def UpperCamelCase_ ( self : Tuple , _A : Dict , _A : Optional[int] , _A : Any , _A : List[Any] , _A : Dict , _A : Union[str, Any] , _A : List[str] ): _UpperCamelCase = self.num_labels _UpperCamelCase = TFDebertaVaForTokenClassification(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) ) def UpperCamelCase_ ( self : Dict , _A : Optional[Any] , _A : Optional[int] , _A : Any , _A : List[str] , _A : str , _A : Optional[int] , _A : str ): _UpperCamelCase = TFDebertaVaForQuestionAnswering(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.start_logits.shape , (self.batch_size, self.seq_length) ) self.parent.assertEqual(result.end_logits.shape , (self.batch_size, self.seq_length) ) def UpperCamelCase_ ( self : Any ): _UpperCamelCase = self.prepare_config_and_inputs() ( ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ) = config_and_inputs _UpperCamelCase = {'''input_ids''': input_ids, '''token_type_ids''': token_type_ids, '''attention_mask''': input_mask} return config, inputs_dict @require_tf class lowerCAmelCase_ ( __lowercase, __lowercase, unittest.TestCase ): UpperCAmelCase = ( ( TFDebertaVaModel, TFDebertaVaForMaskedLM, TFDebertaVaForQuestionAnswering, TFDebertaVaForSequenceClassification, TFDebertaVaForTokenClassification, ) if is_tf_available() else () ) UpperCAmelCase = ( { "feature-extraction": TFDebertaVaModel, "fill-mask": TFDebertaVaForMaskedLM, "question-answering": TFDebertaVaForQuestionAnswering, "text-classification": TFDebertaVaForSequenceClassification, "token-classification": TFDebertaVaForTokenClassification, "zero-shot": TFDebertaVaForSequenceClassification, } if is_tf_available() else {} ) UpperCAmelCase = False UpperCAmelCase = False def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = TFDebertaVaModelTester(self ) _UpperCamelCase = ConfigTester(self , config_class=_A , hidden_size=37 ) def UpperCamelCase_ ( self : Any ): self.config_tester.run_common_tests() def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model(*_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_masked_lm(*_A ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_question_answering(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_sequence_classification(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_token_classification(*_A ) @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = TFDebertaVaModel.from_pretrained('''kamalkraj/deberta-v2-xlarge''' ) self.assertIsNotNone(_A ) @require_tf class lowerCAmelCase_ ( unittest.TestCase ): @unittest.skip(reason='''Model not available yet''' ) def UpperCamelCase_ ( self : List[Any] ): pass @slow def UpperCamelCase_ ( self : int ): _UpperCamelCase = TFDebertaVaModel.from_pretrained('''kamalkraj/deberta-v2-xlarge''' ) _UpperCamelCase = tf.constant([[0, 3_1414, 232, 328, 740, 1140, 1_2695, 69, 4_6078, 1588, 2]] ) _UpperCamelCase = tf.constant([[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]] ) _UpperCamelCase = model(_A , attention_mask=_A )[0] _UpperCamelCase = tf.constant( [[[0.2356, 0.1948, 0.0369], [-0.1063, 0.3586, -0.5152], [-0.6399, -0.0259, -0.2525]]] ) tf.debugging.assert_near(output[:, 1:4, 1:4] , _A , atol=1e-4 )
71
1
from __future__ import annotations import unittest from transformers import BlenderbotConfig, BlenderbotTokenizer, is_tf_available from transformers.testing_utils import require_tf, require_tokenizers, slow from transformers.utils import cached_property from ...test_configuration_common import ConfigTester from ...test_modeling_tf_common import TFModelTesterMixin, ids_tensor from ...test_pipeline_mixin import PipelineTesterMixin if is_tf_available(): import tensorflow as tf from transformers import TFAutoModelForSeqaSeqLM, TFBlenderbotForConditionalGeneration, TFBlenderbotModel @require_tf class lowerCAmelCase_ : UpperCAmelCase = BlenderbotConfig UpperCAmelCase = {} UpperCAmelCase = "gelu" def __init__( self : int , _A : List[Any] , _A : str=13 , _A : Any=7 , _A : Union[str, Any]=True , _A : List[Any]=False , _A : Dict=99 , _A : Optional[Any]=32 , _A : Tuple=2 , _A : List[str]=4 , _A : Union[str, Any]=37 , _A : str=0.1 , _A : List[Any]=0.1 , _A : int=20 , _A : int=2 , _A : Optional[Any]=1 , _A : Union[str, Any]=0 , ): _UpperCamelCase = parent _UpperCamelCase = batch_size _UpperCamelCase = seq_length _UpperCamelCase = is_training _UpperCamelCase = use_labels _UpperCamelCase = vocab_size _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_dropout_prob _UpperCamelCase = attention_probs_dropout_prob _UpperCamelCase = max_position_embeddings _UpperCamelCase = eos_token_id _UpperCamelCase = pad_token_id _UpperCamelCase = bos_token_id def UpperCamelCase_ ( self : str ): _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length - 1] , self.vocab_size ) _UpperCamelCase = tf.expand_dims(tf.constant([self.eos_token_id] * self.batch_size ) , 1 ) _UpperCamelCase = tf.concat([input_ids, eos_tensor] , axis=1 ) _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size ) _UpperCamelCase = self.config_cls( vocab_size=self.vocab_size , d_model=self.hidden_size , encoder_layers=self.num_hidden_layers , decoder_layers=self.num_hidden_layers , encoder_attention_heads=self.num_attention_heads , decoder_attention_heads=self.num_attention_heads , encoder_ffn_dim=self.intermediate_size , decoder_ffn_dim=self.intermediate_size , dropout=self.hidden_dropout_prob , attention_dropout=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , eos_token_ids=[2] , bos_token_id=self.bos_token_id , pad_token_id=self.pad_token_id , decoder_start_token_id=self.pad_token_id , **self.config_updates , ) _UpperCamelCase = prepare_blenderbot_inputs_dict(_A , _A , _A ) return config, inputs_dict def UpperCamelCase_ ( self : Tuple , _A : Any , _A : Optional[int] ): _UpperCamelCase = TFBlenderbotModel(config=_A ).get_decoder() _UpperCamelCase = inputs_dict['''input_ids'''] _UpperCamelCase = input_ids[:1, :] _UpperCamelCase = inputs_dict['''attention_mask'''][:1, :] _UpperCamelCase = inputs_dict['''head_mask'''] _UpperCamelCase = 1 # first forward pass _UpperCamelCase = model(_A , attention_mask=_A , head_mask=_A , use_cache=_A ) _UpperCamelCase , _UpperCamelCase = outputs.to_tuple() # create hypothetical next token and extent to next_input_ids _UpperCamelCase = ids_tensor((self.batch_size, 3) , config.vocab_size ) _UpperCamelCase = tf.cast(ids_tensor((self.batch_size, 3) , 2 ) , tf.inta ) # append to next input_ids and _UpperCamelCase = tf.concat([input_ids, next_tokens] , axis=-1 ) _UpperCamelCase = tf.concat([attention_mask, next_attn_mask] , axis=-1 ) _UpperCamelCase = model(_A , attention_mask=_A )[0] _UpperCamelCase = model(_A , attention_mask=_A , past_key_values=_A )[0] self.parent.assertEqual(next_tokens.shape[1] , output_from_past.shape[1] ) # select random slice _UpperCamelCase = int(ids_tensor((1,) , output_from_past.shape[-1] ) ) _UpperCamelCase = output_from_no_past[:, -3:, random_slice_idx] _UpperCamelCase = output_from_past[:, :, random_slice_idx] # test that outputs are equal for slice tf.debugging.assert_near(_A , _A , rtol=1e-3 ) def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case=None , __snake_case=None , __snake_case=None , __snake_case=None , __snake_case=None , ): if attention_mask is None: _UpperCamelCase = tf.cast(tf.math.not_equal(__snake_case , config.pad_token_id ) , tf.inta ) if decoder_attention_mask is None: _UpperCamelCase = tf.concat( [ tf.ones(decoder_input_ids[:, :1].shape , dtype=tf.inta ), tf.cast(tf.math.not_equal(decoder_input_ids[:, 1:] , config.pad_token_id ) , tf.inta ), ] , axis=-1 , ) if head_mask is None: _UpperCamelCase = tf.ones((config.encoder_layers, config.encoder_attention_heads) ) if decoder_head_mask is None: _UpperCamelCase = tf.ones((config.decoder_layers, config.decoder_attention_heads) ) if cross_attn_head_mask is None: _UpperCamelCase = tf.ones((config.decoder_layers, config.decoder_attention_heads) ) return { "input_ids": input_ids, "decoder_input_ids": decoder_input_ids, "attention_mask": attention_mask, "decoder_attention_mask": decoder_attention_mask, "head_mask": head_mask, "decoder_head_mask": decoder_head_mask, "cross_attn_head_mask": cross_attn_head_mask, } @require_tf class lowerCAmelCase_ ( __lowercase, __lowercase, unittest.TestCase ): UpperCAmelCase = (TFBlenderbotForConditionalGeneration, TFBlenderbotModel) if is_tf_available() else () UpperCAmelCase = (TFBlenderbotForConditionalGeneration,) if is_tf_available() else () UpperCAmelCase = ( { "conversational": TFBlenderbotForConditionalGeneration, "feature-extraction": TFBlenderbotModel, "summarization": TFBlenderbotForConditionalGeneration, "text2text-generation": TFBlenderbotForConditionalGeneration, "translation": TFBlenderbotForConditionalGeneration, } if is_tf_available() else {} ) UpperCAmelCase = True UpperCAmelCase = False UpperCAmelCase = False def UpperCamelCase_ ( self : Any ): _UpperCamelCase = TFBlenderbotModelTester(self ) _UpperCamelCase = ConfigTester(self , config_class=_A ) def UpperCamelCase_ ( self : List[Any] ): self.config_tester.run_common_tests() def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() self.model_tester.check_decoder_model_past_large_inputs(*_A ) @require_tokenizers @require_tf class lowerCAmelCase_ ( unittest.TestCase ): UpperCAmelCase = ["My friends are cool but they eat too many carbs."] UpperCAmelCase = "facebook/blenderbot-400M-distill" @cached_property def UpperCamelCase_ ( self : Any ): return BlenderbotTokenizer.from_pretrained(self.model_name ) @cached_property def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = TFAutoModelForSeqaSeqLM.from_pretrained(self.model_name ) return model @slow def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = self.tokenizer(self.src_text , return_tensors='''tf''' ) _UpperCamelCase = self.model.generate( model_inputs.input_ids , ) _UpperCamelCase = self.tokenizer.batch_decode(generated_ids.numpy() , skip_special_tokens=_A )[0] assert ( generated_words == " That's unfortunate. Are they trying to lose weight or are they just trying to be healthier?" )
71
def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): # Return True if there is node that has not iterated. _UpperCamelCase = [False] * len(__snake_case ) _UpperCamelCase = [] queue.append(__snake_case ) _UpperCamelCase = True while queue: _UpperCamelCase = queue.pop(0 ) for ind in range(len(graph[u] ) ): if visited[ind] is False and graph[u][ind] > 0: queue.append(__snake_case ) _UpperCamelCase = True _UpperCamelCase = u return visited[t] def _snake_case ( __snake_case , __snake_case , __snake_case ): # This array is filled by BFS and to store path _UpperCamelCase = [-1] * (len(__snake_case )) _UpperCamelCase = 0 while bfs(__snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = float('''Inf''' ) _UpperCamelCase = sink while s != source: # Find the minimum value in select path _UpperCamelCase = min(__snake_case , graph[parent[s]][s] ) _UpperCamelCase = parent[s] max_flow += path_flow _UpperCamelCase = sink while v != source: _UpperCamelCase = parent[v] graph[u][v] -= path_flow graph[v][u] += path_flow _UpperCamelCase = parent[v] return max_flow _lowerCAmelCase = [ [0, 16, 13, 0, 0, 0], [0, 0, 10, 12, 0, 0], [0, 4, 0, 0, 14, 0], [0, 0, 9, 0, 0, 20], [0, 0, 0, 7, 0, 4], [0, 0, 0, 0, 0, 0], ] _lowerCAmelCase, _lowerCAmelCase = 0, 5 print(ford_fulkerson(graph, source, sink))
71
1
from __future__ import annotations import collections import tempfile import unittest import numpy as np from transformers.testing_utils import require_tf, require_vision, slow from transformers.utils import is_tf_available, is_vision_available from ...test_modeling_tf_common import floats_tensor, ids_tensor, random_attention_mask from ..bert.test_modeling_tf_bert import TFBertModelTester from ..clip.test_modeling_tf_clip import TFCLIPVisionModelTester from ..deit.test_modeling_tf_deit import TFDeiTModelTester from ..roberta.test_modeling_tf_roberta import TFRobertaModelTester from ..vit.test_modeling_tf_vit import TFViTModelTester if is_tf_available(): from transformers import ( TFBertModel, TFCLIPVisionModel, TFDeiTModel, TFRobertaModel, TFVisionTextDualEncoderModel, TFViTModel, VisionTextDualEncoderConfig, ) if is_vision_available(): from PIL import Image from transformers import VisionTextDualEncoderProcessor def _snake_case ( __snake_case ): if isinstance(__snake_case , collections.abc.Iterable ): return x return (x, x) @require_tf class lowerCAmelCase_ : def UpperCamelCase_ ( self : List[str] , _A : Any , _A : Dict ): pass def UpperCamelCase_ ( self : List[Any] ): pass def UpperCamelCase_ ( self : Optional[Any] ): pass def UpperCamelCase_ ( self : int , _A : Union[str, Any] , _A : Any , _A : Optional[Any] , _A : Optional[Any] , _A : str=None , **_A : Union[str, Any] ): _UpperCamelCase = VisionTextDualEncoderConfig.from_vision_text_configs(_A , _A ) _UpperCamelCase = TFVisionTextDualEncoderModel(_A ) _UpperCamelCase = model(input_ids=_A , pixel_values=_A , attention_mask=_A ) self.assertEqual(output['''text_embeds'''].shape , (input_ids.shape[0], config.projection_dim) ) self.assertEqual(output['''image_embeds'''].shape , (pixel_values.shape[0], config.projection_dim) ) def UpperCamelCase_ ( self : Any , _A : List[Any] , _A : List[Any] , _A : Optional[int] , _A : Union[str, Any] , _A : Tuple=None , **_A : Optional[Any] ): _UpperCamelCase , _UpperCamelCase = self.get_vision_text_model(_A , _A ) _UpperCamelCase = TFVisionTextDualEncoderModel(vision_model=_A , text_model=_A ) _UpperCamelCase = model(input_ids=_A , pixel_values=_A , attention_mask=_A ) self.assertEqual(output['''text_embeds'''].shape , (input_ids.shape[0], model.config.projection_dim) ) self.assertEqual(output['''image_embeds'''].shape , (pixel_values.shape[0], model.config.projection_dim) ) def UpperCamelCase_ ( self : List[Any] , _A : Any , _A : Tuple , _A : List[Any] , _A : Dict , _A : Optional[Any]=None , **_A : Union[str, Any] ): _UpperCamelCase , _UpperCamelCase = self.get_vision_text_model(_A , _A ) _UpperCamelCase = {'''vision_model''': vision_model, '''text_model''': text_model} _UpperCamelCase = TFVisionTextDualEncoderModel.from_vision_text_pretrained(**_A ) _UpperCamelCase = model(input_ids=_A , pixel_values=_A , attention_mask=_A ) self.assertEqual(output['''text_embeds'''].shape , (input_ids.shape[0], model.config.projection_dim) ) self.assertEqual(output['''image_embeds'''].shape , (pixel_values.shape[0], model.config.projection_dim) ) def UpperCamelCase_ ( self : int , _A : List[Any] , _A : List[Any] , _A : Tuple , _A : int , _A : str=None , **_A : Any ): _UpperCamelCase , _UpperCamelCase = self.get_vision_text_model(_A , _A ) _UpperCamelCase = TFVisionTextDualEncoderModel(vision_model=_A , text_model=_A ) _UpperCamelCase = model(input_ids=_A , pixel_values=_A , attention_mask=_A ) _UpperCamelCase = output[0].numpy() with tempfile.TemporaryDirectory() as tmpdirname: model.save_pretrained(_A ) _UpperCamelCase = TFVisionTextDualEncoderModel.from_pretrained(_A ) _UpperCamelCase = model(input_ids=_A , pixel_values=_A , attention_mask=_A ) _UpperCamelCase = after_output[0].numpy() _UpperCamelCase = np.amax(np.abs(out_a - out_a ) ) self.assertLessEqual(_A , 1e-5 ) def UpperCamelCase_ ( self : Dict , _A : int , _A : Optional[Any] , _A : Optional[Any] , _A : Any , _A : Dict=None , **_A : str ): _UpperCamelCase , _UpperCamelCase = self.get_vision_text_model(_A , _A ) _UpperCamelCase = TFVisionTextDualEncoderModel(vision_model=_A , text_model=_A ) _UpperCamelCase = model( input_ids=_A , pixel_values=_A , attention_mask=_A , output_attentions=_A ) _UpperCamelCase = output.vision_model_output.attentions self.assertEqual(len(_A ) , vision_config.num_hidden_layers ) # in ViT, the seq_len equals the number of patches + 1 (we add 1 for the [CLS] token) _UpperCamelCase = to_atuple(vision_model.config.image_size ) _UpperCamelCase = to_atuple(vision_model.config.patch_size ) _UpperCamelCase = (image_size[1] // patch_size[1]) * (image_size[0] // patch_size[0]) _UpperCamelCase = num_patches + 1 self.assertEqual(vision_attentions[0].shape[-3:] , (vision_config.num_attention_heads, seq_len, seq_len) ) _UpperCamelCase = output.text_model_output.attentions self.assertEqual(len(_A ) , text_config.num_hidden_layers ) self.assertEqual( text_attentions[0].shape[-3:] , (text_config.num_attention_heads, input_ids.shape[-1], input_ids.shape[-1]) , ) def UpperCamelCase_ ( self : Union[str, Any] , _A : np.ndarray , _A : np.ndarray , _A : float ): _UpperCamelCase = np.abs((a - b) ).max() self.assertLessEqual(_A , _A , F"""Difference between torch and flax is {diff} (>= {tol}).""" ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.prepare_config_and_inputs() self.check_vision_text_dual_encoder_model(**_A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = self.prepare_config_and_inputs() self.check_model_from_pretrained_configs(**_A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = self.prepare_config_and_inputs() self.check_vision_text_dual_encoder_from_pretrained(**_A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = self.prepare_config_and_inputs() self.check_save_load(**_A ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.prepare_config_and_inputs() self.check_vision_text_output_attention(**_A ) @slow def UpperCamelCase_ ( self : str ): _UpperCamelCase , _UpperCamelCase = self.get_pretrained_model_and_inputs() _UpperCamelCase = model_a(**_A ) _UpperCamelCase = outputs[0].numpy() with tempfile.TemporaryDirectory() as tmp_dirname: model_a.save_pretrained(_A ) _UpperCamelCase = TFVisionTextDualEncoderModel.from_pretrained(_A ) _UpperCamelCase = model_a(**_A ) _UpperCamelCase = after_outputs[0].numpy() _UpperCamelCase = np.amax(np.abs(out_a - out_a ) ) self.assertLessEqual(_A , 1e-5 ) @require_tf class lowerCAmelCase_ ( __lowercase, unittest.TestCase ): def UpperCamelCase_ ( self : str ): _UpperCamelCase = TFVisionTextDualEncoderModel.from_vision_text_pretrained( '''hf-internal-testing/tiny-random-vit''' , '''hf-internal-testing/tiny-random-bert''' ) _UpperCamelCase = 13 _UpperCamelCase = floats_tensor( [ batch_size, model.vision_model.config.num_channels, model.vision_model.config.image_size, model.vision_model.config.image_size, ] ) _UpperCamelCase = ids_tensor([batch_size, 4] , model.text_model.config.vocab_size ) _UpperCamelCase = random_attention_mask([batch_size, 4] ) _UpperCamelCase = {'''pixel_values''': pixel_values, '''input_ids''': input_ids, '''attention_mask''': attention_mask} return model, inputs def UpperCamelCase_ ( self : Any , _A : Optional[int] , _A : Tuple ): _UpperCamelCase = TFViTModel(_A , name='''vision_model''' ) _UpperCamelCase = TFBertModel(_A , name='''text_model''' ) return vision_model, text_model def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = TFViTModelTester(self ) _UpperCamelCase = TFBertModelTester(self ) _UpperCamelCase = vit_model_tester.prepare_config_and_inputs() _UpperCamelCase = bert_model_tester.prepare_config_and_inputs() _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = vision_config_and_inputs ( ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ) = text_config_and_inputs return { "text_config": text_config, "vision_config": vision_config, "pixel_values": pixel_values, "attention_mask": input_mask, "input_ids": input_ids, "text_token_type_ids": token_type_ids, "text_sequence_labels": sequence_labels, "text_token_labels": token_labels, "text_choice_labels": choice_labels, } @require_tf class lowerCAmelCase_ ( __lowercase, unittest.TestCase ): def UpperCamelCase_ ( self : Union[str, Any] ): # DeiT repo doesn't have TF weights, but we don't actually use the weights at all so let's # just reinitialize it. _UpperCamelCase = TFVisionTextDualEncoderModel.from_vision_text_pretrained( '''Rocketknight1/tiny-random-deit-tf''' , '''hf-internal-testing/tiny-random-roberta''' ) _UpperCamelCase = 13 _UpperCamelCase = floats_tensor( [ batch_size, model.vision_model.config.num_channels, model.vision_model.config.image_size, model.vision_model.config.image_size, ] ) _UpperCamelCase = ids_tensor([batch_size, 4] , model.text_model.config.vocab_size ) _UpperCamelCase = random_attention_mask([batch_size, 4] ) _UpperCamelCase = {'''pixel_values''': pixel_values, '''input_ids''': input_ids, '''attention_mask''': attention_mask} return model, inputs def UpperCamelCase_ ( self : Tuple , _A : Optional[int] , _A : Optional[int] , _A : str , _A : Tuple , _A : str=None , **_A : Union[str, Any] ): _UpperCamelCase , _UpperCamelCase = self.get_vision_text_model(_A , _A ) _UpperCamelCase = TFVisionTextDualEncoderModel(vision_model=_A , text_model=_A ) _UpperCamelCase = model( input_ids=_A , pixel_values=_A , attention_mask=_A , output_attentions=_A ) _UpperCamelCase = output.vision_model_output.attentions self.assertEqual(len(_A ) , vision_config.num_hidden_layers ) # in DEiT, the seq_len equals the number of patches + 2 (we add 2 for the [CLS] and distillation tokens) _UpperCamelCase = to_atuple(vision_model.config.image_size ) _UpperCamelCase = to_atuple(vision_model.config.patch_size ) _UpperCamelCase = (image_size[1] // patch_size[1]) * (image_size[0] // patch_size[0]) _UpperCamelCase = num_patches + 2 self.assertEqual(vision_attentions[0].shape[-3:] , (vision_config.num_attention_heads, seq_len, seq_len) ) _UpperCamelCase = output.text_model_output.attentions self.assertEqual(len(_A ) , text_config.num_hidden_layers ) self.assertEqual( text_attentions[0].shape[-3:] , (text_config.num_attention_heads, input_ids.shape[-1], input_ids.shape[-1]) , ) def UpperCamelCase_ ( self : str , _A : Optional[int] , _A : Tuple ): _UpperCamelCase = TFDeiTModel(_A , name='''vision_model''' ) _UpperCamelCase = TFRobertaModel(_A , name='''text_model''' ) return vision_model, text_model def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = TFDeiTModelTester(self ) _UpperCamelCase = TFRobertaModelTester(self ) _UpperCamelCase = vit_model_tester.prepare_config_and_inputs() _UpperCamelCase = bert_model_tester.prepare_config_and_inputs() _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = vision_config_and_inputs ( ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ) = text_config_and_inputs return { "text_config": text_config, "vision_config": vision_config, "pixel_values": pixel_values, "attention_mask": input_mask, "input_ids": input_ids, "text_token_type_ids": token_type_ids, "text_sequence_labels": sequence_labels, "text_token_labels": token_labels, "text_choice_labels": choice_labels, } @require_tf class lowerCAmelCase_ ( __lowercase, unittest.TestCase ): def UpperCamelCase_ ( self : str ): _UpperCamelCase = TFVisionTextDualEncoderModel.from_vision_text_pretrained( '''Rocketknight1/tiny-random-clip-tf''' , '''hf-internal-testing/tiny-random-bert''' ) _UpperCamelCase = 13 _UpperCamelCase = floats_tensor( [ batch_size, model.vision_model.config.num_channels, model.vision_model.config.image_size, model.vision_model.config.image_size, ] ) _UpperCamelCase = ids_tensor([batch_size, 4] , model.text_model.config.vocab_size ) _UpperCamelCase = random_attention_mask([batch_size, 4] ) _UpperCamelCase = {'''pixel_values''': pixel_values, '''input_ids''': input_ids, '''attention_mask''': attention_mask} return model, inputs def UpperCamelCase_ ( self : List[Any] , _A : Any , _A : Tuple ): _UpperCamelCase = TFCLIPVisionModel(_A , name='''vision_model''' ) _UpperCamelCase = TFBertModel(_A , name='''text_model''' ) return vision_model, text_model def UpperCamelCase_ ( self : int ): _UpperCamelCase = TFCLIPVisionModelTester(self ) _UpperCamelCase = TFBertModelTester(self ) _UpperCamelCase = clip_model_tester.prepare_config_and_inputs() _UpperCamelCase = bert_model_tester.prepare_config_and_inputs() _UpperCamelCase , _UpperCamelCase = vision_config_and_inputs ( ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ) = text_config_and_inputs return { "text_config": text_config, "vision_config": vision_config, "pixel_values": pixel_values, "attention_mask": input_mask, "input_ids": input_ids, "text_token_type_ids": token_type_ids, "text_sequence_labels": sequence_labels, "text_token_labels": token_labels, "text_choice_labels": choice_labels, } @require_vision @require_tf class lowerCAmelCase_ ( unittest.TestCase ): @slow def UpperCamelCase_ ( self : int ): _UpperCamelCase = TFVisionTextDualEncoderModel.from_pretrained( '''clip-italian/clip-italian''' , logit_scale_init_value=1.0 , from_pt=_A ) _UpperCamelCase = VisionTextDualEncoderProcessor.from_pretrained('''clip-italian/clip-italian''' ) _UpperCamelCase = Image.open('''./tests/fixtures/tests_samples/COCO/000000039769.png''' ) _UpperCamelCase = processor( text=['''una foto di un gatto''', '''una foto di un cane'''] , images=_A , padding=_A , return_tensors='''np''' ) _UpperCamelCase = model(**_A ) # verify the logits self.assertEqual(outputs.logits_per_image.shape , (inputs.pixel_values.shape[0], inputs.input_ids.shape[0]) ) self.assertEqual( outputs.logits_per_text.shape , (inputs.input_ids.shape[0], inputs.pixel_values.shape[0]) , ) _UpperCamelCase = np.array([[1.228_4727, 0.310_4122]] ) self.assertTrue(np.allclose(outputs.logits_per_image.numpy() , _A , atol=1e-3 ) )
71
from typing import TYPE_CHECKING from ...utils import ( OptionalDependencyNotAvailable, _LazyModule, is_flax_available, is_tf_available, is_torch_available, ) _lowerCAmelCase = {"configuration_unispeech": ["UNISPEECH_PRETRAINED_CONFIG_ARCHIVE_MAP", "UniSpeechConfig"]} try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "UNISPEECH_PRETRAINED_MODEL_ARCHIVE_LIST", "UniSpeechForCTC", "UniSpeechForPreTraining", "UniSpeechForSequenceClassification", "UniSpeechModel", "UniSpeechPreTrainedModel", ] if TYPE_CHECKING: from .configuration_unispeech import UNISPEECH_PRETRAINED_CONFIG_ARCHIVE_MAP, UniSpeechConfig try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_unispeech import ( UNISPEECH_PRETRAINED_MODEL_ARCHIVE_LIST, UniSpeechForCTC, UniSpeechForPreTraining, UniSpeechForSequenceClassification, UniSpeechModel, UniSpeechPreTrainedModel, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
1
from typing import Any, Callable, Dict, List, Optional, Union import torch from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer from diffusers import ( AutoencoderKL, DDIMScheduler, DiffusionPipeline, LMSDiscreteScheduler, PNDMScheduler, StableDiffusionPipeline, UNetaDConditionModel, ) from diffusers.pipelines.stable_diffusion import StableDiffusionPipelineOutput from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker _lowerCAmelCase = "CompVis/stable-diffusion-v1-1" _lowerCAmelCase = "CompVis/stable-diffusion-v1-2" _lowerCAmelCase = "CompVis/stable-diffusion-v1-3" _lowerCAmelCase = "CompVis/stable-diffusion-v1-4" class lowerCAmelCase_ ( __lowercase ): def __init__( self : Dict , _A : AutoencoderKL , _A : CLIPTextModel , _A : CLIPTokenizer , _A : UNetaDConditionModel , _A : Union[DDIMScheduler, PNDMScheduler, LMSDiscreteScheduler] , _A : StableDiffusionSafetyChecker , _A : CLIPImageProcessor , _A : bool = True , ): super()._init_() _UpperCamelCase = StableDiffusionPipeline.from_pretrained(_A ) _UpperCamelCase = StableDiffusionPipeline.from_pretrained(_A ) _UpperCamelCase = StableDiffusionPipeline.from_pretrained(_A ) _UpperCamelCase = StableDiffusionPipeline( vae=_A , text_encoder=_A , tokenizer=_A , unet=_A , scheduler=_A , safety_checker=_A , feature_extractor=_A , requires_safety_checker=_A , ) self.register_modules(pipelinea=self.pipea , pipelinea=self.pipea , pipelinea=self.pipea , pipelinea=self.pipea ) @property def UpperCamelCase_ ( self : Optional[Any] ): return {k: getattr(self , _A ) for k in self.config.keys() if not k.startswith('''_''' )} def UpperCamelCase_ ( self : List[str] , _A : Optional[Union[str, int]] = "auto" ): if slice_size == "auto": # half the attention head size is usually a good trade-off between # speed and memory _UpperCamelCase = self.unet.config.attention_head_dim // 2 self.unet.set_attention_slice(_A ) def UpperCamelCase_ ( self : Any ): self.enable_attention_slicing(_A ) @torch.no_grad() def UpperCamelCase_ ( self : Optional[Any] , _A : Union[str, List[str]] , _A : int = 512 , _A : int = 512 , _A : int = 50 , _A : float = 7.5 , _A : Optional[Union[str, List[str]]] = None , _A : Optional[int] = 1 , _A : float = 0.0 , _A : Optional[torch.Generator] = None , _A : Optional[torch.FloatTensor] = None , _A : Optional[str] = "pil" , _A : bool = True , _A : Optional[Callable[[int, int, torch.FloatTensor], None]] = None , _A : int = 1 , **_A : Dict , ): return self.pipea( prompt=_A , height=_A , width=_A , num_inference_steps=_A , guidance_scale=_A , negative_prompt=_A , num_images_per_prompt=_A , eta=_A , generator=_A , latents=_A , output_type=_A , return_dict=_A , callback=_A , callback_steps=_A , **_A , ) @torch.no_grad() def UpperCamelCase_ ( self : Any , _A : Union[str, List[str]] , _A : int = 512 , _A : int = 512 , _A : int = 50 , _A : float = 7.5 , _A : Optional[Union[str, List[str]]] = None , _A : Optional[int] = 1 , _A : float = 0.0 , _A : Optional[torch.Generator] = None , _A : Optional[torch.FloatTensor] = None , _A : Optional[str] = "pil" , _A : bool = True , _A : Optional[Callable[[int, int, torch.FloatTensor], None]] = None , _A : int = 1 , **_A : Optional[Any] , ): return self.pipea( prompt=_A , height=_A , width=_A , num_inference_steps=_A , guidance_scale=_A , negative_prompt=_A , num_images_per_prompt=_A , eta=_A , generator=_A , latents=_A , output_type=_A , return_dict=_A , callback=_A , callback_steps=_A , **_A , ) @torch.no_grad() def UpperCamelCase_ ( self : List[str] , _A : Union[str, List[str]] , _A : int = 512 , _A : int = 512 , _A : int = 50 , _A : float = 7.5 , _A : Optional[Union[str, List[str]]] = None , _A : Optional[int] = 1 , _A : float = 0.0 , _A : Optional[torch.Generator] = None , _A : Optional[torch.FloatTensor] = None , _A : Optional[str] = "pil" , _A : bool = True , _A : Optional[Callable[[int, int, torch.FloatTensor], None]] = None , _A : int = 1 , **_A : List[Any] , ): return self.pipea( prompt=_A , height=_A , width=_A , num_inference_steps=_A , guidance_scale=_A , negative_prompt=_A , num_images_per_prompt=_A , eta=_A , generator=_A , latents=_A , output_type=_A , return_dict=_A , callback=_A , callback_steps=_A , **_A , ) @torch.no_grad() def UpperCamelCase_ ( self : Dict , _A : Union[str, List[str]] , _A : int = 512 , _A : int = 512 , _A : int = 50 , _A : float = 7.5 , _A : Optional[Union[str, List[str]]] = None , _A : Optional[int] = 1 , _A : float = 0.0 , _A : Optional[torch.Generator] = None , _A : Optional[torch.FloatTensor] = None , _A : Optional[str] = "pil" , _A : bool = True , _A : Optional[Callable[[int, int, torch.FloatTensor], None]] = None , _A : int = 1 , **_A : Optional[int] , ): return self.pipea( prompt=_A , height=_A , width=_A , num_inference_steps=_A , guidance_scale=_A , negative_prompt=_A , num_images_per_prompt=_A , eta=_A , generator=_A , latents=_A , output_type=_A , return_dict=_A , callback=_A , callback_steps=_A , **_A , ) @torch.no_grad() def UpperCamelCase_ ( self : Dict , _A : Union[str, List[str]] , _A : int = 512 , _A : int = 512 , _A : int = 50 , _A : float = 7.5 , _A : Optional[Union[str, List[str]]] = None , _A : Optional[int] = 1 , _A : float = 0.0 , _A : Optional[torch.Generator] = None , _A : Optional[torch.FloatTensor] = None , _A : Optional[str] = "pil" , _A : bool = True , _A : Optional[Callable[[int, int, torch.FloatTensor], None]] = None , _A : int = 1 , **_A : List[Any] , ): _UpperCamelCase = '''cuda''' if torch.cuda.is_available() else '''cpu''' self.to(_A ) # Checks if the height and width are divisible by 8 or not if height % 8 != 0 or width % 8 != 0: raise ValueError(F"""`height` and `width` must be divisible by 8 but are {height} and {width}.""" ) # Get first result from Stable Diffusion Checkpoint v1.1 _UpperCamelCase = self.textaimg_sda_a( prompt=_A , height=_A , width=_A , num_inference_steps=_A , guidance_scale=_A , negative_prompt=_A , num_images_per_prompt=_A , eta=_A , generator=_A , latents=_A , output_type=_A , return_dict=_A , callback=_A , callback_steps=_A , **_A , ) # Get first result from Stable Diffusion Checkpoint v1.2 _UpperCamelCase = self.textaimg_sda_a( prompt=_A , height=_A , width=_A , num_inference_steps=_A , guidance_scale=_A , negative_prompt=_A , num_images_per_prompt=_A , eta=_A , generator=_A , latents=_A , output_type=_A , return_dict=_A , callback=_A , callback_steps=_A , **_A , ) # Get first result from Stable Diffusion Checkpoint v1.3 _UpperCamelCase = self.textaimg_sda_a( prompt=_A , height=_A , width=_A , num_inference_steps=_A , guidance_scale=_A , negative_prompt=_A , num_images_per_prompt=_A , eta=_A , generator=_A , latents=_A , output_type=_A , return_dict=_A , callback=_A , callback_steps=_A , **_A , ) # Get first result from Stable Diffusion Checkpoint v1.4 _UpperCamelCase = self.textaimg_sda_a( prompt=_A , height=_A , width=_A , num_inference_steps=_A , guidance_scale=_A , negative_prompt=_A , num_images_per_prompt=_A , eta=_A , generator=_A , latents=_A , output_type=_A , return_dict=_A , callback=_A , callback_steps=_A , **_A , ) # Get all result images into a single list and pass it via StableDiffusionPipelineOutput for final result return StableDiffusionPipelineOutput([resa[0], resa[0], resa[0], resa[0]] )
71
import json import os import shutil import tempfile import unittest import numpy as np from transformers import BertTokenizerFast from transformers.models.bert.tokenization_bert import VOCAB_FILES_NAMES, BertTokenizer from transformers.testing_utils import require_tokenizers, require_vision from transformers.utils import IMAGE_PROCESSOR_NAME, is_vision_available if is_vision_available(): from PIL import Image from transformers import VisionTextDualEncoderProcessor, ViTImageProcessor @require_tokenizers @require_vision class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = tempfile.mkdtemp() # fmt: off _UpperCamelCase = ['''[UNK]''', '''[CLS]''', '''[SEP]''', '''[PAD]''', '''[MASK]''', '''want''', '''##want''', '''##ed''', '''wa''', '''un''', '''runn''', '''##ing''', ''',''', '''low''', '''lowest'''] # fmt: on _UpperCamelCase = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES['''vocab_file'''] ) with open(self.vocab_file , '''w''' , encoding='''utf-8''' ) as vocab_writer: vocab_writer.write(''''''.join([x + '''\n''' for x in vocab_tokens] ) ) _UpperCamelCase = { '''do_resize''': True, '''size''': {'''height''': 18, '''width''': 18}, '''do_normalize''': True, '''image_mean''': [0.5, 0.5, 0.5], '''image_std''': [0.5, 0.5, 0.5], } _UpperCamelCase = os.path.join(self.tmpdirname , _A ) with open(self.image_processor_file , '''w''' , encoding='''utf-8''' ) as fp: json.dump(_A , _A ) def UpperCamelCase_ ( self : Tuple , **_A : Optional[Any] ): return BertTokenizer.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : List[Any] , **_A : Union[str, Any] ): return ViTImageProcessor.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : int ): shutil.rmtree(self.tmpdirname ) def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = [np.random.randint(255 , size=(3, 30, 400) , dtype=np.uinta )] _UpperCamelCase = [Image.fromarray(np.moveaxis(_A , 0 , -1 ) ) for x in image_inputs] return image_inputs def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = self.get_image_processor() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) processor.save_pretrained(self.tmpdirname ) _UpperCamelCase = VisionTextDualEncoderProcessor.from_pretrained(self.tmpdirname ) self.assertEqual(processor.tokenizer.get_vocab() , tokenizer.get_vocab() ) self.assertIsInstance(processor.tokenizer , (BertTokenizer, BertTokenizerFast) ) self.assertEqual(processor.image_processor.to_json_string() , image_processor.to_json_string() ) self.assertIsInstance(processor.image_processor , _A ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = VisionTextDualEncoderProcessor( tokenizer=self.get_tokenizer() , image_processor=self.get_image_processor() ) processor.save_pretrained(self.tmpdirname ) _UpperCamelCase = self.get_tokenizer(bos_token='''(BOS)''' , eos_token='''(EOS)''' ) _UpperCamelCase = self.get_image_processor(do_normalize=_A , padding_value=1.0 ) _UpperCamelCase = VisionTextDualEncoderProcessor.from_pretrained( self.tmpdirname , bos_token='''(BOS)''' , eos_token='''(EOS)''' , do_normalize=_A , padding_value=1.0 ) self.assertEqual(processor.tokenizer.get_vocab() , tokenizer_add_kwargs.get_vocab() ) self.assertIsInstance(processor.tokenizer , (BertTokenizer, BertTokenizerFast) ) self.assertEqual(processor.image_processor.to_json_string() , image_processor_add_kwargs.to_json_string() ) self.assertIsInstance(processor.image_processor , _A ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = image_processor(_A , return_tensors='''np''' ) _UpperCamelCase = processor(images=_A , return_tensors='''np''' ) for key in input_feat_extract.keys(): self.assertAlmostEqual(input_feat_extract[key].sum() , input_processor[key].sum() , delta=1e-2 ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = processor(text=_A ) _UpperCamelCase = tokenizer(_A ) for key in encoded_tok.keys(): self.assertListEqual(encoded_tok[key] , encoded_processor[key] ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = processor(text=_A , images=_A ) self.assertListEqual(list(inputs.keys() ) , ['''input_ids''', '''token_type_ids''', '''attention_mask''', '''pixel_values'''] ) # test if it raises when no input is passed with self.assertRaises(_A ): processor() def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = [[1, 4, 5, 8, 1, 0, 8], [3, 4, 3, 1, 1, 8, 9]] _UpperCamelCase = processor.batch_decode(_A ) _UpperCamelCase = tokenizer.batch_decode(_A ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = processor(text=_A , images=_A ) self.assertListEqual(list(inputs.keys() ) , processor.model_input_names )
71
1
import argparse import json import os import fairseq import torch from fairseq.data import Dictionary from transformers import ( UniSpeechConfig, UniSpeechForCTC, UniSpeechForPreTraining, WavaVecaFeatureExtractor, WavaVecaPhonemeCTCTokenizer, WavaVecaProcessor, logging, ) logging.set_verbosity_info() _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "post_extract_proj": "feature_projection.projection", "encoder.pos_conv.0": "encoder.pos_conv_embed.conv", "self_attn.k_proj": "encoder.layers.*.attention.k_proj", "self_attn.v_proj": "encoder.layers.*.attention.v_proj", "self_attn.q_proj": "encoder.layers.*.attention.q_proj", "self_attn.out_proj": "encoder.layers.*.attention.out_proj", "self_attn_layer_norm": "encoder.layers.*.layer_norm", "fc1": "encoder.layers.*.feed_forward.intermediate_dense", "fc2": "encoder.layers.*.feed_forward.output_dense", "final_layer_norm": "encoder.layers.*.final_layer_norm", "encoder.layer_norm": "encoder.layer_norm", "w2v_model.layer_norm": "feature_projection.layer_norm", "quantizer.weight_proj": "quantizer.weight_proj", "quantizer.vars": "quantizer.codevectors", "project_q": "project_q", "final_proj": "project_hid", "w2v_encoder.proj": "ctc_proj", "mask_emb": "masked_spec_embed", } _lowerCAmelCase = [ "ctc_proj", "quantizer.weight_proj", "quantizer.codevectors", "project_q", "project_hid", ] def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): for attribute in key.split('''.''' ): if is_finetuned: if attribute in ["quantizer", "project_q", "project_hid"]: # those layers are only relevant for pretraining and should be dropped return if attribute == "ctc_proj": # we should rename `ctc_proj` to `lm_head` for fine-tuned phoneme models _UpperCamelCase = '''lm_head''' _UpperCamelCase = getattr(__snake_case , __snake_case ) if weight_type is not None: _UpperCamelCase = getattr(__snake_case , __snake_case ).shape else: _UpperCamelCase = hf_pointer.shape assert hf_shape == value.shape, ( f"""Shape of hf {key + "." + weight_type if weight_type is not None else ""} is {hf_shape}, but should be""" f""" {value.shape} for {full_name}""" ) if weight_type == "weight": _UpperCamelCase = value elif weight_type == "weight_g": _UpperCamelCase = value elif weight_type == "weight_v": _UpperCamelCase = value elif weight_type == "bias": _UpperCamelCase = value else: _UpperCamelCase = value logger.info(f"""{key + "." + weight_type if weight_type is not None else ""} was initialized from {full_name}.""" ) def _snake_case ( __snake_case , __snake_case , __snake_case ): _UpperCamelCase = [] _UpperCamelCase = fairseq_model.state_dict() _UpperCamelCase = hf_model.unispeech.feature_extractor for name, value in fairseq_dict.items(): _UpperCamelCase = False if "conv_layers" in name: load_conv_layer( __snake_case , __snake_case , __snake_case , __snake_case , hf_model.config.feat_extract_norm == '''group''' , ) _UpperCamelCase = True else: for key, mapped_key in MAPPING.items(): _UpperCamelCase = '''unispeech.''' + mapped_key if mapped_key not in TOP_LEVEL_KEYS else mapped_key if key in name or key.split('''w2v_model.''' )[-1] == name.split('''.''' )[0]: _UpperCamelCase = True if "*" in mapped_key: _UpperCamelCase = name.split(__snake_case )[0].split('''.''' )[-2] _UpperCamelCase = mapped_key.replace('''*''' , __snake_case ) if "weight_g" in name: _UpperCamelCase = '''weight_g''' elif "weight_v" in name: _UpperCamelCase = '''weight_v''' elif "bias" in name: _UpperCamelCase = '''bias''' elif "weight" in name: # TODO: don't match quantizer.weight_proj _UpperCamelCase = '''weight''' else: _UpperCamelCase = None set_recursively(__snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ) continue if not is_used: unused_weights.append(__snake_case ) logger.warning(f"""Unused weights: {unused_weights}""" ) def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = full_name.split('''conv_layers.''' )[-1] _UpperCamelCase = name.split('''.''' ) _UpperCamelCase = int(items[0] ) _UpperCamelCase = int(items[1] ) if type_id == 0: if "bias" in name: assert value.shape == feature_extractor.conv_layers[layer_id].conv.bias.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor.conv_layers[layer_id].conv.bias.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract conv layer {layer_id} was initialized from {full_name}.""" ) elif "weight" in name: assert value.shape == feature_extractor.conv_layers[layer_id].conv.weight.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor.conv_layers[layer_id].conv.weight.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract conv layer {layer_id} was initialized from {full_name}.""" ) elif (type_id == 2 and not use_group_norm) or (type_id == 2 and layer_id == 0 and use_group_norm): if "bias" in name: assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.bias.data.shape, ( f"""{full_name} has size {value.shape}, but {feature_extractor[layer_id].layer_norm.bias.data.shape} was""" " found." ) _UpperCamelCase = value logger.info(f"""Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.""" ) elif "weight" in name: assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.weight.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor[layer_id].layer_norm.weight.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.""" ) else: unused_weights.append(__snake_case ) @torch.no_grad() def _snake_case ( __snake_case , __snake_case , __snake_case=None , __snake_case=None , __snake_case=True ): if config_path is not None: _UpperCamelCase = UniSpeechConfig.from_pretrained(__snake_case ) else: _UpperCamelCase = UniSpeechConfig() if is_finetuned: if dict_path: _UpperCamelCase = Dictionary.load_from_json(__snake_case ) # important change bos & pad token id since CTC symbol is <pad> and # not <s> as in fairseq _UpperCamelCase = target_dict.pad_index _UpperCamelCase = target_dict.bos_index _UpperCamelCase = target_dict.eos_index _UpperCamelCase = len(target_dict.symbols ) _UpperCamelCase = os.path.join(__snake_case , '''vocab.json''' ) if not os.path.isdir(__snake_case ): logger.error('''--pytorch_dump_folder_path ({}) should be a directory'''.format(__snake_case ) ) return os.makedirs(__snake_case , exist_ok=__snake_case ) _UpperCamelCase = target_dict.indices # fairseq has the <pad> and <s> switched _UpperCamelCase = 42 _UpperCamelCase = 43 with open(__snake_case , '''w''' , encoding='''utf-8''' ) as vocab_handle: json.dump(__snake_case , __snake_case ) _UpperCamelCase = WavaVecaPhonemeCTCTokenizer( __snake_case , unk_token=target_dict.unk_word , pad_token=target_dict.pad_word , bos_token=target_dict.bos_word , eos_token=target_dict.eos_word , word_delimiter_token='''|''' , do_lower_case=__snake_case , ) _UpperCamelCase = True if config.feat_extract_norm == '''layer''' else False _UpperCamelCase = WavaVecaFeatureExtractor( feature_size=1 , sampling_rate=16000 , padding_value=0 , do_normalize=__snake_case , return_attention_mask=__snake_case , ) _UpperCamelCase = WavaVecaProcessor(feature_extractor=__snake_case , tokenizer=__snake_case ) processor.save_pretrained(__snake_case ) _UpperCamelCase = UniSpeechForCTC(__snake_case ) else: _UpperCamelCase = UniSpeechForPreTraining(__snake_case ) if is_finetuned: _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = fairseq.checkpoint_utils.load_model_ensemble_and_task( [checkpoint_path] , arg_overrides={'''data''': '''/'''.join(dict_path.split('''/''' )[:-1] ), '''w2v_path''': checkpoint_path} ) else: _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = fairseq.checkpoint_utils.load_model_ensemble_and_task([checkpoint_path] ) _UpperCamelCase = model[0].eval() recursively_load_weights(__snake_case , __snake_case , __snake_case ) hf_unispeech.save_pretrained(__snake_case ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() parser.add_argument("--pytorch_dump_folder_path", default=None, type=str, help="Path to the output PyTorch model.") parser.add_argument("--checkpoint_path", default=None, type=str, help="Path to fairseq checkpoint") parser.add_argument("--dict_path", default=None, type=str, help="Path to dict of fine-tuned model") parser.add_argument("--config_path", default=None, type=str, help="Path to hf config.json of model to convert") parser.add_argument( "--not_finetuned", action="store_true", help="Whether the model to convert is a fine-tuned model or not" ) _lowerCAmelCase = parser.parse_args() convert_unispeech_checkpoint( args.checkpoint_path, args.pytorch_dump_folder_path, args.config_path, args.dict_path, not args.not_finetuned )
71
def _snake_case ( __snake_case , __snake_case , __snake_case ): if n == 0: return 1 elif n % 2 == 1: return (binary_exponentiation(__snake_case , n - 1 , __snake_case ) * a) % mod else: _UpperCamelCase = binary_exponentiation(__snake_case , n / 2 , __snake_case ) return (b * b) % mod # a prime number _lowerCAmelCase = 701 _lowerCAmelCase = 1_000_000_000 _lowerCAmelCase = 10 # using binary exponentiation function, O(log(p)): print((a / b) % p == (a * binary_exponentiation(b, p - 2, p)) % p) print((a / b) % p == (a * b ** (p - 2)) % p)
71
1
def _snake_case ( __snake_case , __snake_case ): if a < 0 or b < 0: raise ValueError('''the value of both inputs must be positive''' ) _UpperCamelCase = str(bin(__snake_case ) )[2:] # remove the leading "0b" _UpperCamelCase = str(bin(__snake_case ) )[2:] # remove the leading "0b" _UpperCamelCase = max(len(__snake_case ) , len(__snake_case ) ) return "0b" + "".join( str(int(char_a != char_b ) ) for char_a, char_b in zip(a_binary.zfill(__snake_case ) , b_binary.zfill(__snake_case ) ) ) if __name__ == "__main__": import doctest doctest.testmod()
71
from math import cos, sin, sqrt, tau from audio_filters.iir_filter import IIRFilter def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 - _cos) / 2 _UpperCamelCase = 1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 + _cos) / 2 _UpperCamelCase = -1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = _sin / 2 _UpperCamelCase = 0 _UpperCamelCase = -ba _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 1 - alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([ba, ba, ba] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = 1 + alpha * big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha * big_a _UpperCamelCase = 1 + alpha / big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha / big_a _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (pmc + aaa) _UpperCamelCase = 2 * big_a * mpc _UpperCamelCase = big_a * (pmc - aaa) _UpperCamelCase = ppmc + aaa _UpperCamelCase = -2 * pmpc _UpperCamelCase = ppmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (ppmc + aaa) _UpperCamelCase = -2 * big_a * pmpc _UpperCamelCase = big_a * (ppmc - aaa) _UpperCamelCase = pmc + aaa _UpperCamelCase = 2 * mpc _UpperCamelCase = pmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt
71
1
def _snake_case ( __snake_case ): _UpperCamelCase , _UpperCamelCase = [], [] while len(__snake_case ) > 1: _UpperCamelCase , _UpperCamelCase = min(__snake_case ), max(__snake_case ) start.append(__snake_case ) end.append(__snake_case ) collection.remove(__snake_case ) collection.remove(__snake_case ) end.reverse() return start + collection + end if __name__ == "__main__": _lowerCAmelCase = input("Enter numbers separated by a comma:\n").strip() _lowerCAmelCase = [int(item) for item in user_input.split(",")] print(*merge_sort(unsorted), sep=",")
71
from ...configuration_utils import PretrainedConfig from ...utils import logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "EleutherAI/gpt-neox-20b": "https://huggingface.co/EleutherAI/gpt-neox-20b/resolve/main/config.json", # See all GPTNeoX models at https://huggingface.co/models?filter=gpt_neox } class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "gpt_neox" def __init__( self : Union[str, Any] , _A : Union[str, Any]=5_0432 , _A : List[Any]=6144 , _A : int=44 , _A : int=64 , _A : Optional[Any]=2_4576 , _A : Any="gelu" , _A : Tuple=0.25 , _A : Union[str, Any]=1_0000 , _A : Tuple=0.0 , _A : Any=0.0 , _A : int=0.1 , _A : List[str]=2048 , _A : Dict=0.02 , _A : Optional[Any]=1e-5 , _A : Tuple=True , _A : List[Any]=0 , _A : Optional[int]=2 , _A : Optional[int]=False , _A : List[Any]=True , _A : Any=None , **_A : Any , ): super().__init__(bos_token_id=_A , eos_token_id=_A , **_A ) _UpperCamelCase = vocab_size _UpperCamelCase = max_position_embeddings _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = rotary_pct _UpperCamelCase = rotary_emb_base _UpperCamelCase = attention_dropout _UpperCamelCase = hidden_dropout _UpperCamelCase = classifier_dropout _UpperCamelCase = initializer_range _UpperCamelCase = layer_norm_eps _UpperCamelCase = use_cache _UpperCamelCase = tie_word_embeddings _UpperCamelCase = use_parallel_residual _UpperCamelCase = rope_scaling self._rope_scaling_validation() if self.hidden_size % self.num_attention_heads != 0: raise ValueError( '''The hidden size is not divisble by the number of attention heads! Make sure to update them!''' ) def UpperCamelCase_ ( self : str ): if self.rope_scaling is None: return if not isinstance(self.rope_scaling , _A ) or len(self.rope_scaling ) != 2: raise ValueError( '''`rope_scaling` must be a dictionary with with two fields, `name` and `factor`, ''' F"""got {self.rope_scaling}""" ) _UpperCamelCase = self.rope_scaling.get('''type''' , _A ) _UpperCamelCase = self.rope_scaling.get('''factor''' , _A ) if rope_scaling_type is None or rope_scaling_type not in ["linear", "dynamic"]: raise ValueError( F"""`rope_scaling`'s name field must be one of ['linear', 'dynamic'], got {rope_scaling_type}""" ) if rope_scaling_factor is None or not isinstance(_A , _A ) or rope_scaling_factor <= 1.0: raise ValueError(F"""`rope_scaling`'s factor field must be an float > 1, got {rope_scaling_factor}""" )
71
1
from __future__ import annotations def _snake_case ( __snake_case ): _UpperCamelCase = 2 _UpperCamelCase = [] while i * i <= n: if n % i: i += 1 else: n //= i factors.append(__snake_case ) if n > 1: factors.append(__snake_case ) return factors if __name__ == "__main__": import doctest doctest.testmod()
71
from ..utils import DummyObject, requires_backends class lowerCAmelCase_ ( metaclass=__lowercase ): UpperCAmelCase = ["keras_nlp"] def __init__( self : Any , *_A : Dict , **_A : List[str] ): requires_backends(self , ['''keras_nlp'''] )
71
1
from sklearn.metrics import mean_squared_error import datasets _lowerCAmelCase = "\\n@article{scikit-learn,\n title={Scikit-learn: Machine Learning in {P}ython},\n author={Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V.\n and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P.\n and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and\n Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E.},\n journal={Journal of Machine Learning Research},\n volume={12},\n pages={2825--2830},\n year={2011}\n}\n" _lowerCAmelCase = "\\nMean Squared Error(MSE) is the average of the square of difference between the predicted\nand actual values.\n" _lowerCAmelCase = "\nArgs:\n predictions: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Estimated target values.\n references: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Ground truth (correct) target values.\n sample_weight: array-like of shape (n_samples,), default=None\n Sample weights.\n multioutput: {\"raw_values\", \"uniform_average\"} or array-like of shape (n_outputs,), default=\"uniform_average\"\n Defines aggregating of multiple output values. Array-like value defines weights used to average errors.\n\n \"raw_values\" : Returns a full set of errors in case of multioutput input.\n\n \"uniform_average\" : Errors of all outputs are averaged with uniform weight.\n\n squared : bool, default=True\n If True returns MSE value, if False returns RMSE (Root Mean Squared Error) value.\n\nReturns:\n mse : mean squared error.\nExamples:\n\n >>> mse_metric = datasets.load_metric(\"mse\")\n >>> predictions = [2.5, 0.0, 2, 8]\n >>> references = [3, -0.5, 2, 7]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.375}\n >>> rmse_result = mse_metric.compute(predictions=predictions, references=references, squared=False)\n >>> print(rmse_result)\n {'mse': 0.6123724356957945}\n\n If you're using multi-dimensional lists, then set the config as follows :\n\n >>> mse_metric = datasets.load_metric(\"mse\", \"multilist\")\n >>> predictions = [[0.5, 1], [-1, 1], [7, -6]]\n >>> references = [[0, 2], [-1, 2], [8, -5]]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.7083333333333334}\n >>> results = mse_metric.compute(predictions=predictions, references=references, multioutput='raw_values')\n >>> print(results) # doctest: +NORMALIZE_WHITESPACE\n {'mse': array([0.41666667, 1. ])}\n" @datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION ) class lowerCAmelCase_ ( datasets.Metric ): def UpperCamelCase_ ( self : Optional[int] ): return datasets.MetricInfo( description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features(self._get_feature_types() ) , reference_urls=[ '''https://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html''' ] , ) def UpperCamelCase_ ( self : Dict ): if self.config_name == "multilist": return { "predictions": datasets.Sequence(datasets.Value('''float''' ) ), "references": datasets.Sequence(datasets.Value('''float''' ) ), } else: return { "predictions": datasets.Value('''float''' ), "references": datasets.Value('''float''' ), } def UpperCamelCase_ ( self : Any , _A : List[Any] , _A : List[str] , _A : Dict=None , _A : List[str]="uniform_average" , _A : int=True ): _UpperCamelCase = mean_squared_error( _A , _A , sample_weight=_A , multioutput=_A , squared=_A ) return {"mse": mse}
71
from typing import Optional, Tuple, Union import tensorflow as tf from ...activations_tf import ACTaFN from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward from ...modeling_tf_outputs import ( TFBaseModelOutputWithNoAttention, TFBaseModelOutputWithPoolingAndNoAttention, TFSequenceClassifierOutput, ) from ...modeling_tf_utils import TFPreTrainedModel, TFSequenceClassificationLoss, keras_serializable, unpack_inputs from ...tf_utils import shape_list from ...utils import logging from .configuration_regnet import RegNetConfig _lowerCAmelCase = logging.get_logger(__name__) # General docstring _lowerCAmelCase = "RegNetConfig" # Base docstring _lowerCAmelCase = "facebook/regnet-y-040" _lowerCAmelCase = [1, 1_088, 7, 7] # Image classification docstring _lowerCAmelCase = "facebook/regnet-y-040" _lowerCAmelCase = "tabby, tabby cat" _lowerCAmelCase = [ "facebook/regnet-y-040", # See all regnet models at https://huggingface.co/models?filter=regnet ] class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : str , _A : int , _A : int = 3 , _A : int = 1 , _A : int = 1 , _A : Optional[str] = "relu" , **_A : Any , ): super().__init__(**_A ) # The padding and conv has been verified in # https://colab.research.google.com/gist/sayakpaul/854bc10eeaf21c9ee2119e0b9f3841a7/scratchpad.ipynb _UpperCamelCase = tf.keras.layers.ZeroPaddingaD(padding=kernel_size // 2 ) _UpperCamelCase = tf.keras.layers.ConvaD( filters=_A , kernel_size=_A , strides=_A , padding='''VALID''' , groups=_A , use_bias=_A , name='''convolution''' , ) _UpperCamelCase = tf.keras.layers.BatchNormalization(epsilon=1e-5 , momentum=0.9 , name='''normalization''' ) _UpperCamelCase = ACTaFN[activation] if activation is not None else tf.identity def UpperCamelCase_ ( self : Any , _A : Any ): _UpperCamelCase = self.convolution(self.padding(_A ) ) _UpperCamelCase = self.normalization(_A ) _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Optional[Any] , _A : RegNetConfig , **_A : Any ): super().__init__(**_A ) _UpperCamelCase = config.num_channels _UpperCamelCase = TFRegNetConvLayer( out_channels=config.embedding_size , kernel_size=3 , stride=2 , activation=config.hidden_act , name='''embedder''' , ) def UpperCamelCase_ ( self : List[str] , _A : Optional[int] ): _UpperCamelCase = shape_list(_A )[1] if tf.executing_eagerly() and num_channels != self.num_channels: raise ValueError( '''Make sure that the channel dimension of the pixel values match with the one set in the configuration.''' ) # When running on CPU, `tf.keras.layers.Conv2D` doesn't support `NCHW` format. # So change the input format from `NCHW` to `NHWC`. # shape = (batch_size, in_height, in_width, in_channels=num_channels) _UpperCamelCase = tf.transpose(_A , perm=(0, 2, 3, 1) ) _UpperCamelCase = self.embedder(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : str , _A : int , _A : int = 2 , **_A : Optional[Any] ): super().__init__(**_A ) _UpperCamelCase = tf.keras.layers.ConvaD( filters=_A , kernel_size=1 , strides=_A , use_bias=_A , name='''convolution''' ) _UpperCamelCase = tf.keras.layers.BatchNormalization(epsilon=1e-5 , momentum=0.9 , name='''normalization''' ) def UpperCamelCase_ ( self : str , _A : tf.Tensor , _A : bool = False ): return self.normalization(self.convolution(_A ) , training=_A ) class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Dict , _A : int , _A : int , **_A : Dict ): super().__init__(**_A ) _UpperCamelCase = tf.keras.layers.GlobalAveragePoolingaD(keepdims=_A , name='''pooler''' ) _UpperCamelCase = [ tf.keras.layers.ConvaD(filters=_A , kernel_size=1 , activation='''relu''' , name='''attention.0''' ), tf.keras.layers.ConvaD(filters=_A , kernel_size=1 , activation='''sigmoid''' , name='''attention.2''' ), ] def UpperCamelCase_ ( self : List[str] , _A : List[Any] ): # [batch_size, h, w, num_channels] -> [batch_size, 1, 1, num_channels] _UpperCamelCase = self.pooler(_A ) for layer_module in self.attention: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = hidden_state * pooled return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , _A : int , _A : int , _A : int = 1 , **_A : str ): super().__init__(**_A ) _UpperCamelCase = in_channels != out_channels or stride != 1 _UpperCamelCase = max(1 , out_channels // config.groups_width ) _UpperCamelCase = ( TFRegNetShortCut(_A , stride=_A , name='''shortcut''' ) if should_apply_shortcut else tf.keras.layers.Activation('''linear''' , name='''shortcut''' ) ) # `self.layers` instead of `self.layer` because that is a reserved argument. _UpperCamelCase = [ TFRegNetConvLayer(_A , kernel_size=1 , activation=config.hidden_act , name='''layer.0''' ), TFRegNetConvLayer( _A , stride=_A , groups=_A , activation=config.hidden_act , name='''layer.1''' ), TFRegNetConvLayer(_A , kernel_size=1 , activation=_A , name='''layer.2''' ), ] _UpperCamelCase = ACTaFN[config.hidden_act] def UpperCamelCase_ ( self : Dict , _A : Tuple ): _UpperCamelCase = hidden_state for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = self.shortcut(_A ) hidden_state += residual _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , _A : int , _A : int , _A : int = 1 , **_A : int ): super().__init__(**_A ) _UpperCamelCase = in_channels != out_channels or stride != 1 _UpperCamelCase = max(1 , out_channels // config.groups_width ) _UpperCamelCase = ( TFRegNetShortCut(_A , stride=_A , name='''shortcut''' ) if should_apply_shortcut else tf.keras.layers.Activation('''linear''' , name='''shortcut''' ) ) _UpperCamelCase = [ TFRegNetConvLayer(_A , kernel_size=1 , activation=config.hidden_act , name='''layer.0''' ), TFRegNetConvLayer( _A , stride=_A , groups=_A , activation=config.hidden_act , name='''layer.1''' ), TFRegNetSELayer(_A , reduced_channels=int(round(in_channels / 4 ) ) , name='''layer.2''' ), TFRegNetConvLayer(_A , kernel_size=1 , activation=_A , name='''layer.3''' ), ] _UpperCamelCase = ACTaFN[config.hidden_act] def UpperCamelCase_ ( self : Tuple , _A : List[Any] ): _UpperCamelCase = hidden_state for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = self.shortcut(_A ) hidden_state += residual _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Tuple , _A : RegNetConfig , _A : int , _A : int , _A : int = 2 , _A : int = 2 , **_A : Union[str, Any] ): super().__init__(**_A ) _UpperCamelCase = TFRegNetXLayer if config.layer_type == '''x''' else TFRegNetYLayer _UpperCamelCase = [ # downsampling is done in the first layer with stride of 2 layer(_A , _A , _A , stride=_A , name='''layers.0''' ), *[layer(_A , _A , _A , name=F"""layers.{i+1}""" ) for i in range(depth - 1 )], ] def UpperCamelCase_ ( self : Union[str, Any] , _A : Optional[int] ): for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , **_A : List[str] ): super().__init__(**_A ) _UpperCamelCase = [] # based on `downsample_in_first_stage`, the first layer of the first stage may or may not downsample the input self.stages.append( TFRegNetStage( _A , config.embedding_size , config.hidden_sizes[0] , stride=2 if config.downsample_in_first_stage else 1 , depth=config.depths[0] , name='''stages.0''' , ) ) _UpperCamelCase = zip(config.hidden_sizes , config.hidden_sizes[1:] ) for i, ((in_channels, out_channels), depth) in enumerate(zip(_A , config.depths[1:] ) ): self.stages.append(TFRegNetStage(_A , _A , _A , depth=_A , name=F"""stages.{i+1}""" ) ) def UpperCamelCase_ ( self : Optional[int] , _A : tf.Tensor , _A : bool = False , _A : bool = True ): _UpperCamelCase = () if output_hidden_states else None for stage_module in self.stages: if output_hidden_states: _UpperCamelCase = hidden_states + (hidden_state,) _UpperCamelCase = stage_module(_A ) if output_hidden_states: _UpperCamelCase = hidden_states + (hidden_state,) if not return_dict: return tuple(v for v in [hidden_state, hidden_states] if v is not None ) return TFBaseModelOutputWithNoAttention(last_hidden_state=_A , hidden_states=_A ) @keras_serializable class lowerCAmelCase_ ( tf.keras.layers.Layer ): UpperCAmelCase = RegNetConfig def __init__( self : int , _A : Tuple , **_A : int ): super().__init__(**_A ) _UpperCamelCase = config _UpperCamelCase = TFRegNetEmbeddings(_A , name='''embedder''' ) _UpperCamelCase = TFRegNetEncoder(_A , name='''encoder''' ) _UpperCamelCase = tf.keras.layers.GlobalAveragePoolingaD(keepdims=_A , name='''pooler''' ) @unpack_inputs def UpperCamelCase_ ( self : Optional[int] , _A : tf.Tensor , _A : Optional[bool] = None , _A : Optional[bool] = None , _A : bool = False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.embedder(_A , training=_A ) _UpperCamelCase = self.encoder( _A , output_hidden_states=_A , return_dict=_A , training=_A ) _UpperCamelCase = encoder_outputs[0] _UpperCamelCase = self.pooler(_A ) # Change to NCHW output format have uniformity in the modules _UpperCamelCase = tf.transpose(_A , perm=(0, 3, 1, 2) ) _UpperCamelCase = tf.transpose(_A , perm=(0, 3, 1, 2) ) # Change the other hidden state outputs to NCHW as well if output_hidden_states: _UpperCamelCase = tuple([tf.transpose(_A , perm=(0, 3, 1, 2) ) for h in encoder_outputs[1]] ) if not return_dict: return (last_hidden_state, pooled_output) + encoder_outputs[1:] return TFBaseModelOutputWithPoolingAndNoAttention( last_hidden_state=_A , pooler_output=_A , hidden_states=hidden_states if output_hidden_states else encoder_outputs.hidden_states , ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = RegNetConfig UpperCAmelCase = "regnet" UpperCAmelCase = "pixel_values" @property def UpperCamelCase_ ( self : Tuple ): return {"pixel_values": tf.TensorSpec(shape=(None, self.config.num_channels, 224, 224) , dtype=tf.floataa )} _lowerCAmelCase = r"\n Parameters:\n This model is a Tensorflow\n [tf.keras.layers.Layer](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Layer) sub-class. Use it as a\n regular Tensorflow Module and refer to the Tensorflow documentation for all matter related to general usage and\n behavior.\n config ([`RegNetConfig`]): Model configuration class with all the parameters of the model.\n Initializing with a config file does not load the weights associated with the model, only the\n configuration. Check out the [`~TFPreTrainedModel.from_pretrained`] method to load the model weights.\n" _lowerCAmelCase = r"\n Args:\n pixel_values (`tf.Tensor` of shape `(batch_size, num_channels, height, width)`):\n Pixel values. Pixel values can be obtained using [`AutoImageProcessor`]. See\n [`ConveNextImageProcessor.__call__`] for details.\n output_hidden_states (`bool`, *optional*):\n Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for\n more detail.\n return_dict (`bool`, *optional*):\n Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.\n" @add_start_docstrings( "The bare RegNet model outputting raw features without any specific head on top.", __lowercase, ) class lowerCAmelCase_ ( __lowercase ): def __init__( self : List[Any] , _A : RegNetConfig , *_A : Optional[int] , **_A : Tuple ): super().__init__(_A , *_A , **_A ) _UpperCamelCase = TFRegNetMainLayer(_A , name='''regnet''' ) @unpack_inputs @add_start_docstrings_to_model_forward(_A ) @add_code_sample_docstrings( checkpoint=_CHECKPOINT_FOR_DOC , output_type=_A , config_class=_CONFIG_FOR_DOC , modality='''vision''' , expected_output=_EXPECTED_OUTPUT_SHAPE , ) def UpperCamelCase_ ( self : Any , _A : tf.Tensor , _A : Optional[bool] = None , _A : Optional[bool] = None , _A : Optional[int]=False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.regnet( pixel_values=_A , output_hidden_states=_A , return_dict=_A , training=_A , ) if not return_dict: return (outputs[0],) + outputs[1:] return TFBaseModelOutputWithPoolingAndNoAttention( last_hidden_state=outputs.last_hidden_state , pooler_output=outputs.pooler_output , hidden_states=outputs.hidden_states , ) @add_start_docstrings( "\n RegNet Model with an image classification head on top (a linear layer on top of the pooled features), e.g. for\n ImageNet.\n ", __lowercase, ) class lowerCAmelCase_ ( __lowercase, __lowercase ): def __init__( self : List[Any] , _A : RegNetConfig , *_A : Any , **_A : int ): super().__init__(_A , *_A , **_A ) _UpperCamelCase = config.num_labels _UpperCamelCase = TFRegNetMainLayer(_A , name='''regnet''' ) # classification head _UpperCamelCase = [ tf.keras.layers.Flatten(), tf.keras.layers.Dense(config.num_labels , name='''classifier.1''' ) if config.num_labels > 0 else tf.identity, ] @unpack_inputs @add_start_docstrings_to_model_forward(_A ) @add_code_sample_docstrings( checkpoint=_IMAGE_CLASS_CHECKPOINT , output_type=_A , config_class=_CONFIG_FOR_DOC , expected_output=_IMAGE_CLASS_EXPECTED_OUTPUT , ) def UpperCamelCase_ ( self : str , _A : tf.Tensor = None , _A : tf.Tensor = None , _A : bool = None , _A : bool = None , _A : Any=False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.regnet( _A , output_hidden_states=_A , return_dict=_A , training=_A ) _UpperCamelCase = outputs.pooler_output if return_dict else outputs[1] _UpperCamelCase = self.classifier[0](_A ) _UpperCamelCase = self.classifier[1](_A ) _UpperCamelCase = None if labels is None else self.hf_compute_loss(labels=_A , logits=_A ) if not return_dict: _UpperCamelCase = (logits,) + outputs[2:] return ((loss,) + output) if loss is not None else output return TFSequenceClassifierOutput(loss=_A , logits=_A , hidden_states=outputs.hidden_states )
71
1
from typing import TYPE_CHECKING from ....utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available _lowerCAmelCase = {"configuration_mmbt": ["MMBTConfig"]} try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = ["MMBTForClassification", "MMBTModel", "ModalEmbeddings"] if TYPE_CHECKING: from .configuration_mmbt import MMBTConfig try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_mmbt import MMBTForClassification, MMBTModel, ModalEmbeddings else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
from sklearn.metrics import mean_squared_error import datasets _lowerCAmelCase = "\\n@article{scikit-learn,\n title={Scikit-learn: Machine Learning in {P}ython},\n author={Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V.\n and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P.\n and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and\n Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E.},\n journal={Journal of Machine Learning Research},\n volume={12},\n pages={2825--2830},\n year={2011}\n}\n" _lowerCAmelCase = "\\nMean Squared Error(MSE) is the average of the square of difference between the predicted\nand actual values.\n" _lowerCAmelCase = "\nArgs:\n predictions: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Estimated target values.\n references: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Ground truth (correct) target values.\n sample_weight: array-like of shape (n_samples,), default=None\n Sample weights.\n multioutput: {\"raw_values\", \"uniform_average\"} or array-like of shape (n_outputs,), default=\"uniform_average\"\n Defines aggregating of multiple output values. Array-like value defines weights used to average errors.\n\n \"raw_values\" : Returns a full set of errors in case of multioutput input.\n\n \"uniform_average\" : Errors of all outputs are averaged with uniform weight.\n\n squared : bool, default=True\n If True returns MSE value, if False returns RMSE (Root Mean Squared Error) value.\n\nReturns:\n mse : mean squared error.\nExamples:\n\n >>> mse_metric = datasets.load_metric(\"mse\")\n >>> predictions = [2.5, 0.0, 2, 8]\n >>> references = [3, -0.5, 2, 7]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.375}\n >>> rmse_result = mse_metric.compute(predictions=predictions, references=references, squared=False)\n >>> print(rmse_result)\n {'mse': 0.6123724356957945}\n\n If you're using multi-dimensional lists, then set the config as follows :\n\n >>> mse_metric = datasets.load_metric(\"mse\", \"multilist\")\n >>> predictions = [[0.5, 1], [-1, 1], [7, -6]]\n >>> references = [[0, 2], [-1, 2], [8, -5]]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.7083333333333334}\n >>> results = mse_metric.compute(predictions=predictions, references=references, multioutput='raw_values')\n >>> print(results) # doctest: +NORMALIZE_WHITESPACE\n {'mse': array([0.41666667, 1. ])}\n" @datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION ) class lowerCAmelCase_ ( datasets.Metric ): def UpperCamelCase_ ( self : Optional[int] ): return datasets.MetricInfo( description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features(self._get_feature_types() ) , reference_urls=[ '''https://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html''' ] , ) def UpperCamelCase_ ( self : Dict ): if self.config_name == "multilist": return { "predictions": datasets.Sequence(datasets.Value('''float''' ) ), "references": datasets.Sequence(datasets.Value('''float''' ) ), } else: return { "predictions": datasets.Value('''float''' ), "references": datasets.Value('''float''' ), } def UpperCamelCase_ ( self : Any , _A : List[Any] , _A : List[str] , _A : Dict=None , _A : List[str]="uniform_average" , _A : int=True ): _UpperCamelCase = mean_squared_error( _A , _A , sample_weight=_A , multioutput=_A , squared=_A ) return {"mse": mse}
71
1
import argparse from transformers import TaConfig, TaForConditionalGeneration, load_tf_weights_in_ta from transformers.utils import logging logging.set_verbosity_info() def _snake_case ( __snake_case , __snake_case , __snake_case ): # Initialise PyTorch model _UpperCamelCase = TaConfig.from_json_file(__snake_case ) print(f"""Building PyTorch model from configuration: {config}""" ) _UpperCamelCase = TaForConditionalGeneration(__snake_case ) # Load weights from tf checkpoint load_tf_weights_in_ta(__snake_case , __snake_case , __snake_case ) # Save pytorch-model print(f"""Save PyTorch model to {pytorch_dump_path}""" ) model.save_pretrained(__snake_case ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() # Required parameters parser.add_argument( "--tf_checkpoint_path", default=None, type=str, required=True, help="Path to the TensorFlow checkpoint path." ) parser.add_argument( "--config_file", default=None, type=str, required=True, help=( "The config json file corresponding to the pre-trained T5 model. \nThis specifies the model architecture." ), ) parser.add_argument( "--pytorch_dump_path", default=None, type=str, required=True, help="Path to the output PyTorch model." ) _lowerCAmelCase = parser.parse_args() convert_tf_checkpoint_to_pytorch(args.tf_checkpoint_path, args.config_file, args.pytorch_dump_path)
71
import os import re import shutil import sys import tempfile import unittest import black _lowerCAmelCase = os.path.abspath(os.path.dirname(os.path.dirname(os.path.dirname(__file__)))) sys.path.append(os.path.join(git_repo_path, "utils")) import check_copies # noqa: E402 # This is the reference code that will be used in the tests. # If DDPMSchedulerOutput is changed in scheduling_ddpm.py, this code needs to be manually updated. _lowerCAmelCase = " \"\"\"\n Output class for the scheduler's step function output.\n\n Args:\n prev_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n Computed sample (x_{t-1}) of previous timestep. `prev_sample` should be used as next model input in the\n denoising loop.\n pred_original_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n The predicted denoised sample (x_{0}) based on the model output from the current timestep.\n `pred_original_sample` can be used to preview progress or for guidance.\n \"\"\"\n\n prev_sample: torch.FloatTensor\n pred_original_sample: Optional[torch.FloatTensor] = None\n" class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = tempfile.mkdtemp() os.makedirs(os.path.join(self.diffusers_dir , '''schedulers/''' ) ) _UpperCamelCase = self.diffusers_dir shutil.copy( os.path.join(_A , '''src/diffusers/schedulers/scheduling_ddpm.py''' ) , os.path.join(self.diffusers_dir , '''schedulers/scheduling_ddpm.py''' ) , ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = '''src/diffusers''' shutil.rmtree(self.diffusers_dir ) def UpperCamelCase_ ( self : Union[str, Any] , _A : Tuple , _A : Optional[Any] , _A : Dict , _A : List[str]=None ): _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + class_code if overwrite_result is not None: _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + overwrite_result _UpperCamelCase = black.Mode(target_versions={black.TargetVersion.PYaa} , line_length=119 ) _UpperCamelCase = black.format_str(_A , mode=_A ) _UpperCamelCase = os.path.join(self.diffusers_dir , '''new_code.py''' ) with open(_A , '''w''' , newline='''\n''' ) as f: f.write(_A ) if overwrite_result is None: self.assertTrue(len(check_copies.is_copy_consistent(_A ) ) == 0 ) else: check_copies.is_copy_consistent(f.name , overwrite=_A ) with open(_A , '''r''' ) as f: self.assertTrue(f.read() , _A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = check_copies.find_code_in_diffusers('''schedulers.scheduling_ddpm.DDPMSchedulerOutput''' ) self.assertEqual(_A , _A ) def UpperCamelCase_ ( self : Optional[Any] ): # Base copy consistency self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , REFERENCE_CODE + '''\n''' , ) # With no empty line at the end self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , _A , ) # Copy consistency with rename self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , re.sub('''DDPM''' , '''Test''' , _A ) , ) # Copy consistency with a really long name _UpperCamelCase = '''TestClassWithAReallyLongNameBecauseSomePeopleLikeThatForSomeReason''' self.check_copy_consistency( F"""# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->{long_class_name}""" , F"""{long_class_name}SchedulerOutput""" , re.sub('''Bert''' , _A , _A ) , ) # Copy consistency with overwrite self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , _A , overwrite_result=re.sub('''DDPM''' , '''Test''' , _A ) , )
71
1
import logging import torch from accelerate import Accelerator from arguments import EvaluationArguments from datasets import load_dataset from torch.utils.data import IterableDataset from torch.utils.data.dataloader import DataLoader from transformers import AutoModelForCausalLM, AutoTokenizer, HfArgumentParser, set_seed class lowerCAmelCase_ ( __lowercase ): def __init__( self : Dict , _A : List[Any] , _A : str , _A : Union[str, Any]=1024 , _A : Tuple=1024 , _A : str=3.6 ): _UpperCamelCase = tokenizer _UpperCamelCase = tokenizer.bos_token_id _UpperCamelCase = dataset _UpperCamelCase = seq_length _UpperCamelCase = seq_length * chars_per_token * num_of_sequences def __iter__( self : Any ): _UpperCamelCase = iter(self.dataset ) _UpperCamelCase = True while more_examples: _UpperCamelCase , _UpperCamelCase = [], 0 while True: if buffer_len >= self.input_characters: break try: buffer.append(next(_A )['''content'''] ) buffer_len += len(buffer[-1] ) except StopIteration: _UpperCamelCase = False break _UpperCamelCase = tokenizer(_A , truncation=_A )['''input_ids'''] _UpperCamelCase = [] for tokenized_input in tokenized_inputs: all_token_ids.extend(tokenized_input + [self.concat_token_id] ) for i in range(0 , len(_A ) , self.seq_length ): _UpperCamelCase = all_token_ids[i : i + self.seq_length] if len(_A ) == self.seq_length: yield torch.tensor(_A ) def _snake_case ( __snake_case ): _UpperCamelCase = {'''streaming''': True} _UpperCamelCase = load_dataset(args.dataset_name , split='''train''' , **__snake_case ) _UpperCamelCase = ConstantLengthDataset(__snake_case , __snake_case , seq_length=args.seq_length ) _UpperCamelCase = DataLoader(__snake_case , batch_size=args.batch_size ) return eval_dataloader def _snake_case ( __snake_case ): model.eval() _UpperCamelCase = [] for step, batch in enumerate(__snake_case ): with torch.no_grad(): _UpperCamelCase = model(__snake_case , labels=__snake_case ) _UpperCamelCase = outputs.loss.repeat(args.batch_size ) losses.append(accelerator.gather(__snake_case ) ) if args.max_eval_steps > 0 and step >= args.max_eval_steps: break _UpperCamelCase = torch.mean(torch.cat(__snake_case ) ) try: _UpperCamelCase = torch.exp(__snake_case ) except OverflowError: _UpperCamelCase = float('''inf''' ) return loss.item(), perplexity.item() # Setup Accelerator _lowerCAmelCase = Accelerator() # Parse configuration _lowerCAmelCase = HfArgumentParser(EvaluationArguments) _lowerCAmelCase = parser.parse_args() set_seed(args.seed) # Logging _lowerCAmelCase = logging.getLogger(__name__) logging.basicConfig( format="%(asctime)s - %(levelname)s - %(name)s - %(message)s", datefmt="%m/%d/%Y %H:%M:%S", level=logging.INFO ) # Load model and tokenizer _lowerCAmelCase = AutoModelForCausalLM.from_pretrained(args.model_ckpt) _lowerCAmelCase = AutoTokenizer.from_pretrained(args.model_ckpt) # Load dataset and dataloader _lowerCAmelCase = create_dataloader(args) # Prepare everything with our `accelerator`. _lowerCAmelCase, _lowerCAmelCase = accelerator.prepare(model, eval_dataloader) # Evaluate and save the last checkpoint logger.info("Evaluating and saving model after training") _lowerCAmelCase, _lowerCAmelCase = evaluate(args) logger.info(f'loss/eval: {eval_loss}, perplexity: {perplexity}')
71
from __future__ import annotations import math class lowerCAmelCase_ : def __init__( self : int , _A : int ): _UpperCamelCase = size # approximate the overall size of segment tree with given value _UpperCamelCase = [0 for i in range(0 , 4 * size )] # create array to store lazy update _UpperCamelCase = [0 for i in range(0 , 4 * size )] _UpperCamelCase = [0 for i in range(0 , 4 * size )] # flag for lazy update def UpperCamelCase_ ( self : str , _A : int ): return idx * 2 def UpperCamelCase_ ( self : Any , _A : int ): return idx * 2 + 1 def UpperCamelCase_ ( self : Union[str, Any] , _A : int , _A : int , _A : int , _A : list[int] ): if left_element == right_element: _UpperCamelCase = a[left_element - 1] else: _UpperCamelCase = (left_element + right_element) // 2 self.build(self.left(_A ) , _A , _A , _A ) self.build(self.right(_A ) , mid + 1 , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) def UpperCamelCase_ ( self : Tuple , _A : int , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return True if left_element >= a and right_element <= b: _UpperCamelCase = val if left_element != right_element: _UpperCamelCase = val _UpperCamelCase = val _UpperCamelCase = True _UpperCamelCase = True return True _UpperCamelCase = (left_element + right_element) // 2 self.update(self.left(_A ) , _A , _A , _A , _A , _A ) self.update(self.right(_A ) , mid + 1 , _A , _A , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) return True def UpperCamelCase_ ( self : Any , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return -math.inf if left_element >= a and right_element <= b: return self.segment_tree[idx] _UpperCamelCase = (left_element + right_element) // 2 _UpperCamelCase = self.query(self.left(_A ) , _A , _A , _A , _A ) _UpperCamelCase = self.query(self.right(_A ) , mid + 1 , _A , _A , _A ) return max(_A , _A ) def __str__( self : Tuple ): return str([self.query(1 , 1 , self.size , _A , _A ) for i in range(1 , self.size + 1 )] ) if __name__ == "__main__": _lowerCAmelCase = [1, 2, -4, 7, 3, -5, 6, 11, -20, 9, 14, 15, 5, 2, -8] _lowerCAmelCase = 15 _lowerCAmelCase = SegmentTree(size) segt.build(1, 1, size, A) print(segt.query(1, 1, size, 4, 6)) print(segt.query(1, 1, size, 7, 11)) print(segt.query(1, 1, size, 7, 12)) segt.update(1, 1, size, 1, 3, 111) print(segt.query(1, 1, size, 1, 15)) segt.update(1, 1, size, 7, 8, 235) print(segt)
71
1
import os import re import shutil import sys import tempfile import unittest import black _lowerCAmelCase = os.path.abspath(os.path.dirname(os.path.dirname(os.path.dirname(__file__)))) sys.path.append(os.path.join(git_repo_path, "utils")) import check_copies # noqa: E402 # This is the reference code that will be used in the tests. # If DDPMSchedulerOutput is changed in scheduling_ddpm.py, this code needs to be manually updated. _lowerCAmelCase = " \"\"\"\n Output class for the scheduler's step function output.\n\n Args:\n prev_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n Computed sample (x_{t-1}) of previous timestep. `prev_sample` should be used as next model input in the\n denoising loop.\n pred_original_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n The predicted denoised sample (x_{0}) based on the model output from the current timestep.\n `pred_original_sample` can be used to preview progress or for guidance.\n \"\"\"\n\n prev_sample: torch.FloatTensor\n pred_original_sample: Optional[torch.FloatTensor] = None\n" class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = tempfile.mkdtemp() os.makedirs(os.path.join(self.diffusers_dir , '''schedulers/''' ) ) _UpperCamelCase = self.diffusers_dir shutil.copy( os.path.join(_A , '''src/diffusers/schedulers/scheduling_ddpm.py''' ) , os.path.join(self.diffusers_dir , '''schedulers/scheduling_ddpm.py''' ) , ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = '''src/diffusers''' shutil.rmtree(self.diffusers_dir ) def UpperCamelCase_ ( self : Union[str, Any] , _A : Tuple , _A : Optional[Any] , _A : Dict , _A : List[str]=None ): _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + class_code if overwrite_result is not None: _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + overwrite_result _UpperCamelCase = black.Mode(target_versions={black.TargetVersion.PYaa} , line_length=119 ) _UpperCamelCase = black.format_str(_A , mode=_A ) _UpperCamelCase = os.path.join(self.diffusers_dir , '''new_code.py''' ) with open(_A , '''w''' , newline='''\n''' ) as f: f.write(_A ) if overwrite_result is None: self.assertTrue(len(check_copies.is_copy_consistent(_A ) ) == 0 ) else: check_copies.is_copy_consistent(f.name , overwrite=_A ) with open(_A , '''r''' ) as f: self.assertTrue(f.read() , _A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = check_copies.find_code_in_diffusers('''schedulers.scheduling_ddpm.DDPMSchedulerOutput''' ) self.assertEqual(_A , _A ) def UpperCamelCase_ ( self : Optional[Any] ): # Base copy consistency self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , REFERENCE_CODE + '''\n''' , ) # With no empty line at the end self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , _A , ) # Copy consistency with rename self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , re.sub('''DDPM''' , '''Test''' , _A ) , ) # Copy consistency with a really long name _UpperCamelCase = '''TestClassWithAReallyLongNameBecauseSomePeopleLikeThatForSomeReason''' self.check_copy_consistency( F"""# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->{long_class_name}""" , F"""{long_class_name}SchedulerOutput""" , re.sub('''Bert''' , _A , _A ) , ) # Copy consistency with overwrite self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , _A , overwrite_result=re.sub('''DDPM''' , '''Test''' , _A ) , )
71
from typing import TYPE_CHECKING from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available _lowerCAmelCase = { "configuration_jukebox": [ "JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP", "JukeboxConfig", "JukeboxPriorConfig", "JukeboxVQVAEConfig", ], "tokenization_jukebox": ["JukeboxTokenizer"], } try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST", "JukeboxModel", "JukeboxPreTrainedModel", "JukeboxVQVAE", "JukeboxPrior", ] if TYPE_CHECKING: from .configuration_jukebox import ( JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP, JukeboxConfig, JukeboxPriorConfig, JukeboxVQVAEConfig, ) from .tokenization_jukebox import JukeboxTokenizer try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_jukebox import ( JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST, JukeboxModel, JukeboxPreTrainedModel, JukeboxPrior, JukeboxVQVAE, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
1
def _snake_case ( __snake_case = 10 ): if not isinstance(__snake_case , __snake_case ) or n < 0: raise ValueError('''Invalid input''' ) _UpperCamelCase = 10**n _UpperCamelCase = 28433 * (pow(2 , 7830457 , __snake_case )) + 1 return str(number % modulus ) if __name__ == "__main__": from doctest import testmod testmod() print(f'{solution(10) = }')
71
import multiprocessing import os from typing import BinaryIO, Optional, Union import fsspec from .. import Dataset, Features, NamedSplit, config from ..formatting import query_table from ..packaged_modules.json.json import Json from ..utils import logging from ..utils.typing import NestedDataStructureLike, PathLike from .abc import AbstractDatasetReader class lowerCAmelCase_ ( __lowercase ): def __init__( self : int , _A : NestedDataStructureLike[PathLike] , _A : Optional[NamedSplit] = None , _A : Optional[Features] = None , _A : str = None , _A : bool = False , _A : bool = False , _A : Optional[str] = None , _A : Optional[int] = None , **_A : str , ): super().__init__( _A , split=_A , features=_A , cache_dir=_A , keep_in_memory=_A , streaming=_A , num_proc=_A , **_A , ) _UpperCamelCase = field _UpperCamelCase = path_or_paths if isinstance(_A , _A ) else {self.split: path_or_paths} _UpperCamelCase = Json( cache_dir=_A , data_files=_A , features=_A , field=_A , **_A , ) def UpperCamelCase_ ( self : List[str] ): # Build iterable dataset if self.streaming: _UpperCamelCase = self.builder.as_streaming_dataset(split=self.split ) # Build regular (map-style) dataset else: _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None self.builder.download_and_prepare( download_config=_A , download_mode=_A , verification_mode=_A , base_path=_A , num_proc=self.num_proc , ) _UpperCamelCase = self.builder.as_dataset( split=self.split , verification_mode=_A , in_memory=self.keep_in_memory ) return dataset class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : Dataset , _A : Union[PathLike, BinaryIO] , _A : Optional[int] = None , _A : Optional[int] = None , **_A : List[str] , ): if num_proc is not None and num_proc <= 0: raise ValueError(F"""num_proc {num_proc} must be an integer > 0.""" ) _UpperCamelCase = dataset _UpperCamelCase = path_or_buf _UpperCamelCase = batch_size if batch_size else config.DEFAULT_MAX_BATCH_SIZE _UpperCamelCase = num_proc _UpperCamelCase = '''utf-8''' _UpperCamelCase = to_json_kwargs def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = self.to_json_kwargs.pop('''path_or_buf''' , _A ) _UpperCamelCase = self.to_json_kwargs.pop('''orient''' , '''records''' ) _UpperCamelCase = self.to_json_kwargs.pop('''lines''' , True if orient == '''records''' else False ) _UpperCamelCase = self.to_json_kwargs.pop('''index''' , False if orient in ['''split''', '''table'''] else True ) _UpperCamelCase = self.to_json_kwargs.pop('''compression''' , _A ) if compression not in [None, "infer", "gzip", "bz2", "xz"]: raise NotImplementedError(F"""`datasets` currently does not support {compression} compression""" ) if isinstance(self.path_or_buf , (str, bytes, os.PathLike) ): with fsspec.open(self.path_or_buf , '''wb''' , compression=_A ) as buffer: _UpperCamelCase = self._write(file_obj=_A , orient=_A , lines=_A , index=_A , **self.to_json_kwargs ) else: if compression: raise NotImplementedError( F"""The compression parameter is not supported when writing to a buffer, but compression={compression}""" ''' was passed. Please provide a local path instead.''' ) _UpperCamelCase = self._write( file_obj=self.path_or_buf , orient=_A , lines=_A , index=_A , **self.to_json_kwargs ) return written def UpperCamelCase_ ( self : Any , _A : Optional[Any] ): _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = args _UpperCamelCase = query_table( table=self.dataset.data , key=slice(_A , offset + self.batch_size ) , indices=self.dataset._indices , ) _UpperCamelCase = batch.to_pandas().to_json( path_or_buf=_A , orient=_A , lines=_A , index=_A , **_A ) if not json_str.endswith('''\n''' ): json_str += "\n" return json_str.encode(self.encoding ) def UpperCamelCase_ ( self : int , _A : BinaryIO , _A : Dict , _A : Optional[Any] , _A : Dict , **_A : str , ): _UpperCamelCase = 0 if self.num_proc is None or self.num_proc == 1: for offset in logging.tqdm( range(0 , len(self.dataset ) , self.batch_size ) , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating json from Arrow format''' , ): _UpperCamelCase = self._batch_json((offset, orient, lines, index, to_json_kwargs) ) written += file_obj.write(_A ) else: _UpperCamelCase , _UpperCamelCase = len(self.dataset ), self.batch_size with multiprocessing.Pool(self.num_proc ) as pool: for json_str in logging.tqdm( pool.imap( self._batch_json , [(offset, orient, lines, index, to_json_kwargs) for offset in range(0 , _A , _A )] , ) , total=(num_rows // batch_size) + 1 if num_rows % batch_size else num_rows // batch_size , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating json from Arrow format''' , ): written += file_obj.write(_A ) return written
71
1
from __future__ import annotations def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , ): _UpperCamelCase = len(__snake_case ) # If row is equal to the size of the board it means there are a queen in each row in # the current board (possible_board) if row == n: # We convert the variable possible_board that looks like this: [1, 3, 0, 2] to # this: ['. Q . . ', '. . . Q ', 'Q . . . ', '. . Q . '] boards.append(['''. ''' * i + '''Q ''' + '''. ''' * (n - 1 - i) for i in possible_board] ) return # We iterate each column in the row to find all possible results in each row for col in range(__snake_case ): # We apply that we learned previously. First we check that in the current board # (possible_board) there are not other same value because if there is it means # that there are a collision in vertical. Then we apply the two formulas we # learned before: # # 45º: y - x = b or 45: row - col = b # 135º: y + x = b or row + col = b. # # And we verify if the results of this two formulas not exist in their variables # respectively. (diagonal_right_collisions, diagonal_left_collisions) # # If any or these are True it means there is a collision so we continue to the # next value in the for loop. if ( col in possible_board or row - col in diagonal_right_collisions or row + col in diagonal_left_collisions ): continue # If it is False we call dfs function again and we update the inputs depth_first_search( [*possible_board, col] , [*diagonal_right_collisions, row - col] , [*diagonal_left_collisions, row + col] , __snake_case , __snake_case , ) def _snake_case ( __snake_case ): _UpperCamelCase = [] depth_first_search([] , [] , [] , __snake_case , __snake_case ) # Print all the boards for board in boards: for column in board: print(__snake_case ) print('''''' ) print(len(__snake_case ) , '''solutions were found.''' ) if __name__ == "__main__": import doctest doctest.testmod() n_queens_solution(4)
71
import enum import warnings from .. import MODEL_FOR_CAUSAL_LM_MAPPING, TF_MODEL_FOR_CAUSAL_LM_MAPPING from ..utils import add_end_docstrings, is_tf_available from .base import PIPELINE_INIT_ARGS, Pipeline if is_tf_available(): import tensorflow as tf class lowerCAmelCase_ ( enum.Enum ): UpperCAmelCase = 0 UpperCAmelCase = 1 UpperCAmelCase = 2 @add_end_docstrings(__lowercase ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "\n In 1991, the remains of Russian Tsar Nicholas II and his family (except for Alexei and Maria) are discovered. The\n voice of Nicholas's young son, Tsarevich Alexei Nikolaevich, narrates the remainder of the story. 1883 Western\n Siberia, a young Grigori Rasputin is asked by his father and a group of men to perform magic. Rasputin has a vision\n and denounces one of the men as a horse thief. Although his father initially slaps him for making such an\n accusation, Rasputin watches as the man is chased outside and beaten. Twenty years later, Rasputin sees a vision of\n the Virgin Mary, prompting him to become a priest. Rasputin quickly becomes famous, with people, even a bishop,\n begging for his blessing. <eod> </s> <eos>\n " def __init__( self : Tuple , *_A : List[str] , **_A : str ): super().__init__(*_A , **_A ) self.check_model_type( TF_MODEL_FOR_CAUSAL_LM_MAPPING if self.framework == '''tf''' else MODEL_FOR_CAUSAL_LM_MAPPING ) if "prefix" not in self._preprocess_params: # This is very specific. The logic is quite complex and needs to be done # as a "default". # It also defines both some preprocess_kwargs and generate_kwargs # which is why we cannot put them in their respective methods. _UpperCamelCase = None if self.model.config.prefix is not None: _UpperCamelCase = self.model.config.prefix if prefix is None and self.model.__class__.__name__ in [ "XLNetLMHeadModel", "TransfoXLLMHeadModel", "TFXLNetLMHeadModel", "TFTransfoXLLMHeadModel", ]: # For XLNet and TransformerXL we add an article to the prompt to give more state to the model. _UpperCamelCase = self.XL_PREFIX if prefix is not None: # Recalculate some generate_kwargs linked to prefix. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = self._sanitize_parameters(prefix=_A , **self._forward_params ) _UpperCamelCase = {**self._preprocess_params, **preprocess_params} _UpperCamelCase = {**self._forward_params, **forward_params} def UpperCamelCase_ ( self : Dict , _A : Optional[int]=None , _A : Any=None , _A : Optional[int]=None , _A : List[str]=None , _A : List[Any]=None , _A : int=None , _A : Tuple=None , _A : Optional[Any]=None , **_A : Optional[int] , ): _UpperCamelCase = {} if prefix is not None: _UpperCamelCase = prefix if prefix: _UpperCamelCase = self.tokenizer( _A , padding=_A , add_special_tokens=_A , return_tensors=self.framework ) _UpperCamelCase = prefix_inputs['''input_ids'''].shape[-1] if handle_long_generation is not None: if handle_long_generation not in {"hole"}: raise ValueError( F"""{handle_long_generation} is not a valid value for `handle_long_generation` parameter expected""" ''' [None, \'hole\']''' ) _UpperCamelCase = handle_long_generation preprocess_params.update(_A ) _UpperCamelCase = generate_kwargs _UpperCamelCase = {} if return_full_text is not None and return_type is None: if return_text is not None: raise ValueError('''`return_text` is mutually exclusive with `return_full_text`''' ) if return_tensors is not None: raise ValueError('''`return_full_text` is mutually exclusive with `return_tensors`''' ) _UpperCamelCase = ReturnType.FULL_TEXT if return_full_text else ReturnType.NEW_TEXT if return_tensors is not None and return_type is None: if return_text is not None: raise ValueError('''`return_text` is mutually exclusive with `return_tensors`''' ) _UpperCamelCase = ReturnType.TENSORS if return_type is not None: _UpperCamelCase = return_type if clean_up_tokenization_spaces is not None: _UpperCamelCase = clean_up_tokenization_spaces if stop_sequence is not None: _UpperCamelCase = self.tokenizer.encode(_A , add_special_tokens=_A ) if len(_A ) > 1: warnings.warn( '''Stopping on a multiple token sequence is not yet supported on transformers. The first token of''' ''' the stop sequence will be used as the stop sequence string in the interim.''' ) _UpperCamelCase = stop_sequence_ids[0] return preprocess_params, forward_params, postprocess_params def UpperCamelCase_ ( self : int , *_A : Union[str, Any] , **_A : Union[str, Any] ): # Parse arguments if self.model.__class__.__name__ in ["TransfoXLLMHeadModel"]: kwargs.update({'''add_space_before_punct_symbol''': True} ) return super()._parse_and_tokenize(*_A , **_A ) def __call__( self : List[str] , _A : str , **_A : Any ): return super().__call__(_A , **_A ) def UpperCamelCase_ ( self : Optional[Any] , _A : List[str] , _A : int="" , _A : Optional[Any]=None , **_A : Optional[Any] ): _UpperCamelCase = self.tokenizer( prefix + prompt_text , padding=_A , add_special_tokens=_A , return_tensors=self.framework ) _UpperCamelCase = prompt_text if handle_long_generation == "hole": _UpperCamelCase = inputs['''input_ids'''].shape[-1] if "max_new_tokens" in generate_kwargs: _UpperCamelCase = generate_kwargs['''max_new_tokens'''] else: _UpperCamelCase = generate_kwargs.get('''max_length''' , self.model.config.max_length ) - cur_len if new_tokens < 0: raise ValueError('''We cannot infer how many new tokens are expected''' ) if cur_len + new_tokens > self.tokenizer.model_max_length: _UpperCamelCase = self.tokenizer.model_max_length - new_tokens if keep_length <= 0: raise ValueError( '''We cannot use `hole` to handle this generation the number of desired tokens exceeds the''' ''' models max length''' ) _UpperCamelCase = inputs['''input_ids'''][:, -keep_length:] if "attention_mask" in inputs: _UpperCamelCase = inputs['''attention_mask'''][:, -keep_length:] return inputs def UpperCamelCase_ ( self : Dict , _A : Optional[int] , **_A : str ): _UpperCamelCase = model_inputs['''input_ids'''] _UpperCamelCase = model_inputs.get('''attention_mask''' , _A ) # Allow empty prompts if input_ids.shape[1] == 0: _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = 1 else: _UpperCamelCase = input_ids.shape[0] _UpperCamelCase = model_inputs.pop('''prompt_text''' ) # If there is a prefix, we may need to adjust the generation length. Do so without permanently modifying # generate_kwargs, as some of the parameterization may come from the initialization of the pipeline. _UpperCamelCase = generate_kwargs.pop('''prefix_length''' , 0 ) if prefix_length > 0: _UpperCamelCase = '''max_new_tokens''' in generate_kwargs or ( '''generation_config''' in generate_kwargs and generate_kwargs['''generation_config'''].max_new_tokens is not None ) if not has_max_new_tokens: _UpperCamelCase = generate_kwargs.get('''max_length''' ) or self.model.config.max_length generate_kwargs["max_length"] += prefix_length _UpperCamelCase = '''min_new_tokens''' in generate_kwargs or ( '''generation_config''' in generate_kwargs and generate_kwargs['''generation_config'''].min_new_tokens is not None ) if not has_min_new_tokens and "min_length" in generate_kwargs: generate_kwargs["min_length"] += prefix_length # BS x SL _UpperCamelCase = self.model.generate(input_ids=_A , attention_mask=_A , **_A ) _UpperCamelCase = generated_sequence.shape[0] if self.framework == "pt": _UpperCamelCase = generated_sequence.reshape(_A , out_b // in_b , *generated_sequence.shape[1:] ) elif self.framework == "tf": _UpperCamelCase = tf.reshape(_A , (in_b, out_b // in_b, *generated_sequence.shape[1:]) ) return {"generated_sequence": generated_sequence, "input_ids": input_ids, "prompt_text": prompt_text} def UpperCamelCase_ ( self : List[str] , _A : Dict , _A : Optional[Any]=ReturnType.FULL_TEXT , _A : Dict=True ): _UpperCamelCase = model_outputs['''generated_sequence'''][0] _UpperCamelCase = model_outputs['''input_ids'''] _UpperCamelCase = model_outputs['''prompt_text'''] _UpperCamelCase = generated_sequence.numpy().tolist() _UpperCamelCase = [] for sequence in generated_sequence: if return_type == ReturnType.TENSORS: _UpperCamelCase = {'''generated_token_ids''': sequence} elif return_type in {ReturnType.NEW_TEXT, ReturnType.FULL_TEXT}: # Decode text _UpperCamelCase = self.tokenizer.decode( _A , skip_special_tokens=_A , clean_up_tokenization_spaces=_A , ) # Remove PADDING prompt of the sequence if XLNet or Transfo-XL model is used if input_ids is None: _UpperCamelCase = 0 else: _UpperCamelCase = len( self.tokenizer.decode( input_ids[0] , skip_special_tokens=_A , clean_up_tokenization_spaces=_A , ) ) if return_type == ReturnType.FULL_TEXT: _UpperCamelCase = prompt_text + text[prompt_length:] else: _UpperCamelCase = text[prompt_length:] _UpperCamelCase = {'''generated_text''': all_text} records.append(_A ) return records
71
1
def _snake_case ( __snake_case ): def merge(__snake_case , __snake_case ) -> list: def _merge(): while left and right: yield (left if left[0] <= right[0] else right).pop(0 ) yield from left yield from right return list(_merge() ) if len(__snake_case ) <= 1: return collection _UpperCamelCase = len(__snake_case ) // 2 return merge(merge_sort(collection[:mid] ) , merge_sort(collection[mid:] ) ) if __name__ == "__main__": import doctest doctest.testmod() _lowerCAmelCase = input("Enter numbers separated by a comma:\n").strip() _lowerCAmelCase = [int(item) for item in user_input.split(",")] print(*merge_sort(unsorted), sep=",")
71
import unittest from transformers import is_torch_available from transformers.testing_utils import require_sentencepiece, require_tokenizers, require_torch, slow, torch_device if is_torch_available(): from transformers import AutoModelForSeqaSeqLM, AutoTokenizer @require_torch @require_sentencepiece @require_tokenizers class lowerCAmelCase_ ( unittest.TestCase ): @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = AutoModelForSeqaSeqLM.from_pretrained('''google/mt5-small''' , return_dict=_A ).to(_A ) _UpperCamelCase = AutoTokenizer.from_pretrained('''google/mt5-small''' ) _UpperCamelCase = tokenizer('''Hello there''' , return_tensors='''pt''' ).input_ids _UpperCamelCase = tokenizer('''Hi I am''' , return_tensors='''pt''' ).input_ids _UpperCamelCase = model(input_ids.to(_A ) , labels=labels.to(_A ) ).loss _UpperCamelCase = -(labels.shape[-1] * loss.item()) _UpperCamelCase = -84.9127 self.assertTrue(abs(mtf_score - EXPECTED_SCORE ) < 1e-4 )
71
1
import argparse import os import torch from diffusers import ( CMStochasticIterativeScheduler, ConsistencyModelPipeline, UNetaDModel, ) _lowerCAmelCase = { "sample_size": 32, "in_channels": 3, "out_channels": 3, "layers_per_block": 2, "num_class_embeds": 1_000, "block_out_channels": [32, 64], "attention_head_dim": 8, "down_block_types": [ "ResnetDownsampleBlock2D", "AttnDownBlock2D", ], "up_block_types": [ "AttnUpBlock2D", "ResnetUpsampleBlock2D", ], "resnet_time_scale_shift": "scale_shift", "upsample_type": "resnet", "downsample_type": "resnet", } _lowerCAmelCase = { "sample_size": 64, "in_channels": 3, "out_channels": 3, "layers_per_block": 3, "num_class_embeds": 1_000, "block_out_channels": [192, 192 * 2, 192 * 3, 192 * 4], "attention_head_dim": 64, "down_block_types": [ "ResnetDownsampleBlock2D", "AttnDownBlock2D", "AttnDownBlock2D", "AttnDownBlock2D", ], "up_block_types": [ "AttnUpBlock2D", "AttnUpBlock2D", "AttnUpBlock2D", "ResnetUpsampleBlock2D", ], "resnet_time_scale_shift": "scale_shift", "upsample_type": "resnet", "downsample_type": "resnet", } _lowerCAmelCase = { "sample_size": 256, "in_channels": 3, "out_channels": 3, "layers_per_block": 2, "num_class_embeds": None, "block_out_channels": [256, 256, 256 * 2, 256 * 2, 256 * 4, 256 * 4], "attention_head_dim": 64, "down_block_types": [ "ResnetDownsampleBlock2D", "ResnetDownsampleBlock2D", "ResnetDownsampleBlock2D", "AttnDownBlock2D", "AttnDownBlock2D", "AttnDownBlock2D", ], "up_block_types": [ "AttnUpBlock2D", "AttnUpBlock2D", "AttnUpBlock2D", "ResnetUpsampleBlock2D", "ResnetUpsampleBlock2D", "ResnetUpsampleBlock2D", ], "resnet_time_scale_shift": "default", "upsample_type": "resnet", "downsample_type": "resnet", } _lowerCAmelCase = { "num_train_timesteps": 40, "sigma_min": 0.002, "sigma_max": 80.0, } _lowerCAmelCase = { "num_train_timesteps": 201, "sigma_min": 0.002, "sigma_max": 80.0, } _lowerCAmelCase = { "num_train_timesteps": 151, "sigma_min": 0.002, "sigma_max": 80.0, } def _snake_case ( __snake_case ): if isinstance(__snake_case , __snake_case ): return v if v.lower() in ("yes", "true", "t", "y", "1"): return True elif v.lower() in ("no", "false", "f", "n", "0"): return False else: raise argparse.ArgumentTypeError('''boolean value expected''' ) def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case=False ): _UpperCamelCase = checkpoint[f"""{old_prefix}.in_layers.0.weight"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.in_layers.0.bias"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.in_layers.2.weight"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.in_layers.2.bias"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.emb_layers.1.weight"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.emb_layers.1.bias"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.out_layers.0.weight"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.out_layers.0.bias"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.out_layers.3.weight"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.out_layers.3.bias"""] if has_skip: _UpperCamelCase = checkpoint[f"""{old_prefix}.skip_connection.weight"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.skip_connection.bias"""] return new_checkpoint def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case=None ): _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = checkpoint[f"""{old_prefix}.qkv.weight"""].chunk(3 , dim=0 ) _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = checkpoint[f"""{old_prefix}.qkv.bias"""].chunk(3 , dim=0 ) _UpperCamelCase = checkpoint[f"""{old_prefix}.norm.weight"""] _UpperCamelCase = checkpoint[f"""{old_prefix}.norm.bias"""] _UpperCamelCase = weight_q.squeeze(-1 ).squeeze(-1 ) _UpperCamelCase = bias_q.squeeze(-1 ).squeeze(-1 ) _UpperCamelCase = weight_k.squeeze(-1 ).squeeze(-1 ) _UpperCamelCase = bias_k.squeeze(-1 ).squeeze(-1 ) _UpperCamelCase = weight_v.squeeze(-1 ).squeeze(-1 ) _UpperCamelCase = bias_v.squeeze(-1 ).squeeze(-1 ) _UpperCamelCase = ( checkpoint[f"""{old_prefix}.proj_out.weight"""].squeeze(-1 ).squeeze(-1 ) ) _UpperCamelCase = checkpoint[f"""{old_prefix}.proj_out.bias"""].squeeze(-1 ).squeeze(-1 ) return new_checkpoint def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = torch.load(__snake_case , map_location='''cpu''' ) _UpperCamelCase = {} _UpperCamelCase = checkpoint['''time_embed.0.weight'''] _UpperCamelCase = checkpoint['''time_embed.0.bias'''] _UpperCamelCase = checkpoint['''time_embed.2.weight'''] _UpperCamelCase = checkpoint['''time_embed.2.bias'''] if unet_config["num_class_embeds"] is not None: _UpperCamelCase = checkpoint['''label_emb.weight'''] _UpperCamelCase = checkpoint['''input_blocks.0.0.weight'''] _UpperCamelCase = checkpoint['''input_blocks.0.0.bias'''] _UpperCamelCase = unet_config['''down_block_types'''] _UpperCamelCase = unet_config['''layers_per_block'''] _UpperCamelCase = unet_config['''attention_head_dim'''] _UpperCamelCase = unet_config['''block_out_channels'''] _UpperCamelCase = 1 _UpperCamelCase = channels_list[0] for i, layer_type in enumerate(__snake_case ): _UpperCamelCase = channels_list[i] _UpperCamelCase = current_channels != prev_channels if layer_type == "ResnetDownsampleBlock2D": for j in range(__snake_case ): _UpperCamelCase = f"""down_blocks.{i}.resnets.{j}""" _UpperCamelCase = f"""input_blocks.{current_layer}.0""" _UpperCamelCase = True if j == 0 and downsample_block_has_skip else False _UpperCamelCase = convert_resnet(__snake_case , __snake_case , __snake_case , __snake_case , has_skip=__snake_case ) current_layer += 1 elif layer_type == "AttnDownBlock2D": for j in range(__snake_case ): _UpperCamelCase = f"""down_blocks.{i}.resnets.{j}""" _UpperCamelCase = f"""input_blocks.{current_layer}.0""" _UpperCamelCase = True if j == 0 and downsample_block_has_skip else False _UpperCamelCase = convert_resnet(__snake_case , __snake_case , __snake_case , __snake_case , has_skip=__snake_case ) _UpperCamelCase = f"""down_blocks.{i}.attentions.{j}""" _UpperCamelCase = f"""input_blocks.{current_layer}.1""" _UpperCamelCase = convert_attention( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ) current_layer += 1 if i != len(__snake_case ) - 1: _UpperCamelCase = f"""down_blocks.{i}.downsamplers.0""" _UpperCamelCase = f"""input_blocks.{current_layer}.0""" _UpperCamelCase = convert_resnet(__snake_case , __snake_case , __snake_case , __snake_case ) current_layer += 1 _UpperCamelCase = current_channels # hardcoded the mid-block for now _UpperCamelCase = '''mid_block.resnets.0''' _UpperCamelCase = '''middle_block.0''' _UpperCamelCase = convert_resnet(__snake_case , __snake_case , __snake_case , __snake_case ) _UpperCamelCase = '''mid_block.attentions.0''' _UpperCamelCase = '''middle_block.1''' _UpperCamelCase = convert_attention(__snake_case , __snake_case , __snake_case , __snake_case , __snake_case ) _UpperCamelCase = '''mid_block.resnets.1''' _UpperCamelCase = '''middle_block.2''' _UpperCamelCase = convert_resnet(__snake_case , __snake_case , __snake_case , __snake_case ) _UpperCamelCase = 0 _UpperCamelCase = unet_config['''up_block_types'''] for i, layer_type in enumerate(__snake_case ): if layer_type == "ResnetUpsampleBlock2D": for j in range(layers_per_block + 1 ): _UpperCamelCase = f"""up_blocks.{i}.resnets.{j}""" _UpperCamelCase = f"""output_blocks.{current_layer}.0""" _UpperCamelCase = convert_resnet(__snake_case , __snake_case , __snake_case , __snake_case , has_skip=__snake_case ) current_layer += 1 if i != len(__snake_case ) - 1: _UpperCamelCase = f"""up_blocks.{i}.upsamplers.0""" _UpperCamelCase = f"""output_blocks.{current_layer-1}.1""" _UpperCamelCase = convert_resnet(__snake_case , __snake_case , __snake_case , __snake_case ) elif layer_type == "AttnUpBlock2D": for j in range(layers_per_block + 1 ): _UpperCamelCase = f"""up_blocks.{i}.resnets.{j}""" _UpperCamelCase = f"""output_blocks.{current_layer}.0""" _UpperCamelCase = convert_resnet(__snake_case , __snake_case , __snake_case , __snake_case , has_skip=__snake_case ) _UpperCamelCase = f"""up_blocks.{i}.attentions.{j}""" _UpperCamelCase = f"""output_blocks.{current_layer}.1""" _UpperCamelCase = convert_attention( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ) current_layer += 1 if i != len(__snake_case ) - 1: _UpperCamelCase = f"""up_blocks.{i}.upsamplers.0""" _UpperCamelCase = f"""output_blocks.{current_layer-1}.2""" _UpperCamelCase = convert_resnet(__snake_case , __snake_case , __snake_case , __snake_case ) _UpperCamelCase = checkpoint['''out.0.weight'''] _UpperCamelCase = checkpoint['''out.0.bias'''] _UpperCamelCase = checkpoint['''out.2.weight'''] _UpperCamelCase = checkpoint['''out.2.bias'''] return new_checkpoint if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() parser.add_argument("--unet_path", default=None, type=str, required=True, help="Path to the unet.pt to convert.") parser.add_argument( "--dump_path", default=None, type=str, required=True, help="Path to output the converted UNet model." ) parser.add_argument("--class_cond", default=True, type=str, help="Whether the model is class-conditional.") _lowerCAmelCase = parser.parse_args() _lowerCAmelCase = strabool(args.class_cond) _lowerCAmelCase = os.path.basename(args.unet_path) print(f'Checkpoint: {ckpt_name}') # Get U-Net config if "imagenet64" in ckpt_name: _lowerCAmelCase = IMAGENET_64_UNET_CONFIG elif "256" in ckpt_name and (("bedroom" in ckpt_name) or ("cat" in ckpt_name)): _lowerCAmelCase = LSUN_256_UNET_CONFIG elif "test" in ckpt_name: _lowerCAmelCase = TEST_UNET_CONFIG else: raise ValueError(f'Checkpoint type {ckpt_name} is not currently supported.') if not args.class_cond: _lowerCAmelCase = None _lowerCAmelCase = con_pt_to_diffuser(args.unet_path, unet_config) _lowerCAmelCase = UNetaDModel(**unet_config) image_unet.load_state_dict(converted_unet_ckpt) # Get scheduler config if "cd" in ckpt_name or "test" in ckpt_name: _lowerCAmelCase = CD_SCHEDULER_CONFIG elif "ct" in ckpt_name and "imagenet64" in ckpt_name: _lowerCAmelCase = CT_IMAGENET_64_SCHEDULER_CONFIG elif "ct" in ckpt_name and "256" in ckpt_name and (("bedroom" in ckpt_name) or ("cat" in ckpt_name)): _lowerCAmelCase = CT_LSUN_256_SCHEDULER_CONFIG else: raise ValueError(f'Checkpoint type {ckpt_name} is not currently supported.') _lowerCAmelCase = CMStochasticIterativeScheduler(**scheduler_config) _lowerCAmelCase = ConsistencyModelPipeline(unet=image_unet, scheduler=cm_scheduler) consistency_model.save_pretrained(args.dump_path)
71
import logging import os import sys from dataclasses import dataclass, field from typing import Optional from seqaseq_trainer import SeqaSeqTrainer from seqaseq_training_args import SeqaSeqTrainingArguments import transformers from transformers import ( AutoConfig, AutoModelForSeqaSeqLM, AutoTokenizer, HfArgumentParser, MBartTokenizer, MBartTokenizerFast, set_seed, ) from transformers.trainer_utils import EvaluationStrategy, is_main_process from transformers.training_args import ParallelMode from utils import ( SeqaSeqDataCollator, SeqaSeqDataset, assert_all_frozen, build_compute_metrics_fn, check_output_dir, freeze_embeds, freeze_params, lmap, save_json, use_task_specific_params, write_txt_file, ) _lowerCAmelCase = logging.getLogger(__name__) @dataclass class lowerCAmelCase_ : UpperCAmelCase = field( metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Pretrained config name or path if not the same as model_name"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"}, ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Whether tp freeze the encoder."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Whether to freeze the embeddings."} ) @dataclass class lowerCAmelCase_ : UpperCAmelCase = field( metadata={"help": "The input data dir. Should contain the .tsv files (or other data files) for the task."} ) UpperCAmelCase = field( default="summarization", metadata={"help": "Task name, summarization (or summarization_{dataset} for pegasus) or translation"}, ) UpperCAmelCase = field( default=1024, metadata={ "help": ( "The maximum total input sequence length after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field( default=128, metadata={ "help": ( "The maximum total sequence length for target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field( default=142, metadata={ "help": ( "The maximum total sequence length for validation target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded. " "This argument is also used to override the ``max_length`` param of ``model.generate``, which is used " "during ``evaluate`` and ``predict``." ) }, ) UpperCAmelCase = field( default=142, metadata={ "help": ( "The maximum total sequence length for test target text after tokenization. Sequences longer " "than this will be truncated, sequences shorter will be padded." ) }, ) UpperCAmelCase = field(default=-1, metadata={"help": "# training examples. -1 means use all."} ) UpperCAmelCase = field(default=-1, metadata={"help": "# validation examples. -1 means use all."} ) UpperCAmelCase = field(default=-1, metadata={"help": "# test examples. -1 means use all."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Source language id for translation."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "Target language id for translation."} ) UpperCAmelCase = field(default=__lowercase, metadata={"help": "# num_beams to use for evaluation."} ) UpperCAmelCase = field( default=__lowercase, metadata={"help": "If only pad tokens should be ignored. This assumes that `config.pad_token_id` is defined."}, ) def _snake_case ( __snake_case , __snake_case , __snake_case ): logger.info(f"""***** {split} metrics *****""" ) for key in sorted(metrics.keys() ): logger.info(f""" {key} = {metrics[key]}""" ) save_json(__snake_case , os.path.join(__snake_case , f"""{split}_results.json""" ) ) def _snake_case ( ): # See all possible arguments in src/transformers/training_args.py # or by passing the --help flag to this script. # We now keep distinct sets of args, for a cleaner separation of concerns. _UpperCamelCase = HfArgumentParser((ModelArguments, DataTrainingArguments, SeqaSeqTrainingArguments) ) if len(sys.argv ) == 2 and sys.argv[1].endswith('''.json''' ): # If we pass only one argument to the script and it's the path to a json file, # let's parse it to get our arguments. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1] ) ) else: _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = parser.parse_args_into_dataclasses() check_output_dir(__snake_case ) # Setup logging logging.basicConfig( format='''%(asctime)s - %(levelname)s - %(name)s - %(message)s''' , datefmt='''%m/%d/%Y %H:%M:%S''' , level=logging.INFO if training_args.local_rank in [-1, 0] else logging.WARN , ) logger.warning( '''Process rank: %s, device: %s, n_gpu: %s, distributed training: %s, 16-bits training: %s''' , training_args.local_rank , training_args.device , training_args.n_gpu , bool(training_args.parallel_mode == ParallelMode.DISTRIBUTED ) , training_args.fpaa , ) transformers.utils.logging.enable_default_handler() transformers.utils.logging.enable_explicit_format() # Set the verbosity to info of the Transformers logger (on main process only): if is_main_process(training_args.local_rank ): transformers.utils.logging.set_verbosity_info() logger.info('''Training/evaluation parameters %s''' , __snake_case ) # Set seed set_seed(training_args.seed ) # Load pretrained model and tokenizer # # Distributed training: # The .from_pretrained methods guarantee that only one local process can concurrently # download model & vocab. _UpperCamelCase = AutoConfig.from_pretrained( model_args.config_name if model_args.config_name else model_args.model_name_or_path , cache_dir=model_args.cache_dir , ) _UpperCamelCase = ('''encoder_layerdrop''', '''decoder_layerdrop''', '''dropout''', '''attention_dropout''') for p in extra_model_params: if getattr(__snake_case , __snake_case , __snake_case ): assert hasattr(__snake_case , __snake_case ), f"""({config.__class__.__name__}) doesn't have a `{p}` attribute""" setattr(__snake_case , __snake_case , getattr(__snake_case , __snake_case ) ) _UpperCamelCase = AutoTokenizer.from_pretrained( model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path , cache_dir=model_args.cache_dir , ) _UpperCamelCase = AutoModelForSeqaSeqLM.from_pretrained( model_args.model_name_or_path , from_tf='''.ckpt''' in model_args.model_name_or_path , config=__snake_case , cache_dir=model_args.cache_dir , ) # use task specific params use_task_specific_params(__snake_case , data_args.task ) # set num_beams for evaluation if data_args.eval_beams is None: _UpperCamelCase = model.config.num_beams # set decoder_start_token_id for MBart if model.config.decoder_start_token_id is None and isinstance(__snake_case , (MBartTokenizer, MBartTokenizerFast) ): assert ( data_args.tgt_lang is not None and data_args.src_lang is not None ), "mBart requires --tgt_lang and --src_lang" if isinstance(__snake_case , __snake_case ): _UpperCamelCase = tokenizer.lang_code_to_id[data_args.tgt_lang] else: _UpperCamelCase = tokenizer.convert_tokens_to_ids(data_args.tgt_lang ) if model_args.freeze_embeds: freeze_embeds(__snake_case ) if model_args.freeze_encoder: freeze_params(model.get_encoder() ) assert_all_frozen(model.get_encoder() ) _UpperCamelCase = SeqaSeqDataset # Get datasets _UpperCamelCase = ( dataset_class( __snake_case , type_path='''train''' , data_dir=data_args.data_dir , n_obs=data_args.n_train , max_target_length=data_args.max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_train else None ) _UpperCamelCase = ( dataset_class( __snake_case , type_path='''val''' , data_dir=data_args.data_dir , n_obs=data_args.n_val , max_target_length=data_args.val_max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_eval or training_args.evaluation_strategy != EvaluationStrategy.NO else None ) _UpperCamelCase = ( dataset_class( __snake_case , type_path='''test''' , data_dir=data_args.data_dir , n_obs=data_args.n_test , max_target_length=data_args.test_max_target_length , max_source_length=data_args.max_source_length , prefix=model.config.prefix or '''''' , ) if training_args.do_predict else None ) # Initialize our Trainer _UpperCamelCase = ( build_compute_metrics_fn(data_args.task , __snake_case ) if training_args.predict_with_generate else None ) _UpperCamelCase = SeqaSeqTrainer( model=__snake_case , args=__snake_case , data_args=__snake_case , train_dataset=__snake_case , eval_dataset=__snake_case , data_collator=SeqaSeqDataCollator( __snake_case , __snake_case , model.config.decoder_start_token_id , training_args.tpu_num_cores ) , compute_metrics=__snake_case , tokenizer=__snake_case , ) _UpperCamelCase = {} # Training if training_args.do_train: logger.info('''*** Train ***''' ) _UpperCamelCase = trainer.train( model_path=model_args.model_name_or_path if os.path.isdir(model_args.model_name_or_path ) else None ) _UpperCamelCase = train_result.metrics _UpperCamelCase = data_args.n_train trainer.save_model() # this also saves the tokenizer if trainer.is_world_process_zero(): handle_metrics('''train''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) # Need to save the state, since Trainer.save_model saves only the tokenizer with the model trainer.state.save_to_json(os.path.join(training_args.output_dir , '''trainer_state.json''' ) ) # For convenience, we also re-save the tokenizer to the same directory, # so that you can share your model easily on huggingface.co/models =) tokenizer.save_pretrained(training_args.output_dir ) # Evaluation if training_args.do_eval: logger.info('''*** Evaluate ***''' ) _UpperCamelCase = trainer.evaluate(metric_key_prefix='''val''' ) _UpperCamelCase = data_args.n_val _UpperCamelCase = round(metrics['''val_loss'''] , 4 ) if trainer.is_world_process_zero(): handle_metrics('''val''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) if training_args.do_predict: logger.info('''*** Predict ***''' ) _UpperCamelCase = trainer.predict(test_dataset=__snake_case , metric_key_prefix='''test''' ) _UpperCamelCase = test_output.metrics _UpperCamelCase = data_args.n_test if trainer.is_world_process_zero(): _UpperCamelCase = round(metrics['''test_loss'''] , 4 ) handle_metrics('''test''' , __snake_case , training_args.output_dir ) all_metrics.update(__snake_case ) if training_args.predict_with_generate: _UpperCamelCase = tokenizer.batch_decode( test_output.predictions , skip_special_tokens=__snake_case , clean_up_tokenization_spaces=__snake_case ) _UpperCamelCase = lmap(str.strip , __snake_case ) write_txt_file(__snake_case , os.path.join(training_args.output_dir , '''test_generations.txt''' ) ) if trainer.is_world_process_zero(): save_json(__snake_case , os.path.join(training_args.output_dir , '''all_results.json''' ) ) return all_metrics def _snake_case ( __snake_case ): # For xla_spawn (TPUs) main() if __name__ == "__main__": main()
71
1
from unittest.mock import patch import pyspark from datasets.packaged_modules.spark.spark import ( Spark, SparkExamplesIterable, _generate_iterable_examples, ) from ..utils import ( require_dill_gt_0_3_2, require_not_windows, ) def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = [] for part_id in partition_order: _UpperCamelCase = df.where(f"""SPARK_PARTITION_ID() = {part_id}""" ).collect() for row_idx, row in enumerate(__snake_case ): expected_row_ids_and_row_dicts.append((f"""{part_id}_{row_idx}""", row.asDict()) ) return expected_row_ids_and_row_dicts @require_not_windows @require_dill_gt_0_3_2 def _snake_case ( ): _UpperCamelCase = pyspark.sql.SparkSession.builder.master('''local[*]''' ).appName('''pyspark''' ).getOrCreate() _UpperCamelCase = spark.range(100 ).repartition(1 ) _UpperCamelCase = Spark(__snake_case ) # The id ints will be converted to Pyarrow int64s, so each row will be 8 bytes. Setting a max_shard_size of 16 means # that each partition can hold 2 rows. spark_builder._repartition_df_if_needed(max_shard_size=16 ) # Given that the dataframe has 100 rows and each partition has 2 rows, we expect 50 partitions. assert spark_builder.df.rdd.getNumPartitions() == 50 @require_not_windows @require_dill_gt_0_3_2 def _snake_case ( ): _UpperCamelCase = pyspark.sql.SparkSession.builder.master('''local[*]''' ).appName('''pyspark''' ).getOrCreate() _UpperCamelCase = spark.range(10 ).repartition(2 ) _UpperCamelCase = [1, 0] _UpperCamelCase = _generate_iterable_examples(__snake_case , __snake_case ) # Reverse the partitions. _UpperCamelCase = _get_expected_row_ids_and_row_dicts_for_partition_order(__snake_case , __snake_case ) for i, (row_id, row_dict) in enumerate(generate_fn() ): _UpperCamelCase , _UpperCamelCase = expected_row_ids_and_row_dicts[i] assert row_id == expected_row_id assert row_dict == expected_row_dict @require_not_windows @require_dill_gt_0_3_2 def _snake_case ( ): _UpperCamelCase = pyspark.sql.SparkSession.builder.master('''local[*]''' ).appName('''pyspark''' ).getOrCreate() _UpperCamelCase = spark.range(10 ).repartition(1 ) _UpperCamelCase = SparkExamplesIterable(__snake_case ) assert it.n_shards == 1 for i, (row_id, row_dict) in enumerate(__snake_case ): assert row_id == f"""0_{i}""" assert row_dict == {"id": i} @require_not_windows @require_dill_gt_0_3_2 def _snake_case ( ): _UpperCamelCase = pyspark.sql.SparkSession.builder.master('''local[*]''' ).appName('''pyspark''' ).getOrCreate() _UpperCamelCase = spark.range(30 ).repartition(3 ) # Mock the generator so that shuffle reverses the partition indices. with patch('''numpy.random.Generator''' ) as generator_mock: _UpperCamelCase = lambda __snake_case : x.reverse() _UpperCamelCase = _get_expected_row_ids_and_row_dicts_for_partition_order(__snake_case , [2, 1, 0] ) _UpperCamelCase = SparkExamplesIterable(__snake_case ).shuffle_data_sources(__snake_case ) assert shuffled_it.n_shards == 3 for i, (row_id, row_dict) in enumerate(__snake_case ): _UpperCamelCase , _UpperCamelCase = expected_row_ids_and_row_dicts[i] assert row_id == expected_row_id assert row_dict == expected_row_dict @require_not_windows @require_dill_gt_0_3_2 def _snake_case ( ): _UpperCamelCase = pyspark.sql.SparkSession.builder.master('''local[*]''' ).appName('''pyspark''' ).getOrCreate() _UpperCamelCase = spark.range(20 ).repartition(4 ) # Partitions 0 and 2 _UpperCamelCase = SparkExamplesIterable(__snake_case ).shard_data_sources(worker_id=0 , num_workers=2 ) assert shard_it_a.n_shards == 2 _UpperCamelCase = _get_expected_row_ids_and_row_dicts_for_partition_order(__snake_case , [0, 2] ) for i, (row_id, row_dict) in enumerate(__snake_case ): _UpperCamelCase , _UpperCamelCase = expected_row_ids_and_row_dicts_a[i] assert row_id == expected_row_id assert row_dict == expected_row_dict # Partitions 1 and 3 _UpperCamelCase = SparkExamplesIterable(__snake_case ).shard_data_sources(worker_id=1 , num_workers=2 ) assert shard_it_a.n_shards == 2 _UpperCamelCase = _get_expected_row_ids_and_row_dicts_for_partition_order(__snake_case , [1, 3] ) for i, (row_id, row_dict) in enumerate(__snake_case ): _UpperCamelCase , _UpperCamelCase = expected_row_ids_and_row_dicts_a[i] assert row_id == expected_row_id assert row_dict == expected_row_dict @require_not_windows @require_dill_gt_0_3_2 def _snake_case ( ): _UpperCamelCase = pyspark.sql.SparkSession.builder.master('''local[*]''' ).appName('''pyspark''' ).getOrCreate() _UpperCamelCase = spark.range(100 ).repartition(1 ) _UpperCamelCase = Spark(__snake_case ) # Choose a small max_shard_size for maximum partitioning. spark_builder._repartition_df_if_needed(max_shard_size=1 ) # The new number of partitions should not be greater than the number of rows. assert spark_builder.df.rdd.getNumPartitions() == 100
71
from __future__ import annotations import typing from collections import Counter def _snake_case ( __snake_case ): _UpperCamelCase = Counter() for base in range(1 , max_perimeter + 1 ): for perpendicular in range(__snake_case , max_perimeter + 1 ): _UpperCamelCase = (base * base + perpendicular * perpendicular) ** 0.5 if hypotenuse == int(__snake_case ): _UpperCamelCase = int(base + perpendicular + hypotenuse ) if perimeter > max_perimeter: continue triplets[perimeter] += 1 return triplets def _snake_case ( __snake_case = 1000 ): _UpperCamelCase = pythagorean_triple(__snake_case ) return triplets.most_common(1 )[0][0] if __name__ == "__main__": print(f'Perimeter {solution()} has maximum solutions')
71
1
import re from typing import Callable, List, Optional, Union import tensorflow as tf try: from tensorflow.keras.optimizers.legacy import Adam except ImportError: from tensorflow.keras.optimizers import Adam class lowerCAmelCase_ ( tf.keras.optimizers.schedules.LearningRateSchedule ): def __init__( self : int , _A : float , _A : Callable , _A : int , _A : float = 1.0 , _A : str = None , ): super().__init__() _UpperCamelCase = initial_learning_rate _UpperCamelCase = warmup_steps _UpperCamelCase = power _UpperCamelCase = decay_schedule_fn _UpperCamelCase = name def __call__( self : Dict , _A : Dict ): with tf.name_scope(self.name or '''WarmUp''' ) as name: # Implements polynomial warmup. i.e., if global_step < warmup_steps, the # learning rate will be `global_step/num_warmup_steps * init_lr`. _UpperCamelCase = tf.cast(_A , tf.floataa ) _UpperCamelCase = tf.cast(self.warmup_steps , tf.floataa ) _UpperCamelCase = global_step_float / warmup_steps_float _UpperCamelCase = self.initial_learning_rate * tf.math.pow(_A , self.power ) return tf.cond( global_step_float < warmup_steps_float , lambda: warmup_learning_rate , lambda: self.decay_schedule_fn(step - self.warmup_steps ) , name=_A , ) def UpperCamelCase_ ( self : str ): return { "initial_learning_rate": self.initial_learning_rate, "decay_schedule_fn": self.decay_schedule_fn, "warmup_steps": self.warmup_steps, "power": self.power, "name": self.name, } def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 0.0 , __snake_case = 0.9 , __snake_case = 0.999 , __snake_case = 1E-8 , __snake_case = None , __snake_case = None , __snake_case = 0.0 , __snake_case = 1.0 , __snake_case = None , ): _UpperCamelCase = tf.keras.optimizers.schedules.PolynomialDecay( initial_learning_rate=__snake_case , decay_steps=num_train_steps - num_warmup_steps , end_learning_rate=init_lr * min_lr_ratio , power=__snake_case , ) if num_warmup_steps: _UpperCamelCase = WarmUp( initial_learning_rate=__snake_case , decay_schedule_fn=__snake_case , warmup_steps=__snake_case , ) if weight_decay_rate > 0.0: _UpperCamelCase = AdamWeightDecay( learning_rate=__snake_case , weight_decay_rate=__snake_case , beta_a=__snake_case , beta_a=__snake_case , epsilon=__snake_case , clipnorm=__snake_case , global_clipnorm=__snake_case , exclude_from_weight_decay=['''LayerNorm''', '''layer_norm''', '''bias'''] , include_in_weight_decay=__snake_case , ) else: _UpperCamelCase = tf.keras.optimizers.Adam( learning_rate=__snake_case , beta_a=__snake_case , beta_a=__snake_case , epsilon=__snake_case , clipnorm=__snake_case , global_clipnorm=__snake_case , ) # We return the optimizer and the LR scheduler in order to better track the # evolution of the LR independently of the optimizer. return optimizer, lr_schedule class lowerCAmelCase_ ( __lowercase ): def __init__( self : Optional[Any] , _A : Union[float, tf.keras.optimizers.schedules.LearningRateSchedule] = 0.001 , _A : float = 0.9 , _A : float = 0.999 , _A : float = 1e-7 , _A : bool = False , _A : float = 0.0 , _A : Optional[List[str]] = None , _A : Optional[List[str]] = None , _A : str = "AdamWeightDecay" , **_A : List[Any] , ): super().__init__(_A , _A , _A , _A , _A , _A , **_A ) _UpperCamelCase = weight_decay_rate _UpperCamelCase = include_in_weight_decay _UpperCamelCase = exclude_from_weight_decay @classmethod def UpperCamelCase_ ( cls : Dict , _A : Optional[int] ): _UpperCamelCase = {'''WarmUp''': WarmUp} return super(_A , cls ).from_config(_A , custom_objects=_A ) def UpperCamelCase_ ( self : Dict , _A : Optional[int] , _A : Optional[Any] , _A : str ): super(_A , self )._prepare_local(_A , _A , _A ) _UpperCamelCase = tf.constant( self.weight_decay_rate , name='''adam_weight_decay_rate''' ) def UpperCamelCase_ ( self : Tuple , _A : Tuple , _A : List[Any] , _A : Optional[Any] ): _UpperCamelCase = self._do_use_weight_decay(var.name ) if do_decay: return var.assign_sub( learning_rate * var * apply_state[(var.device, var.dtype.base_dtype)]['''weight_decay_rate'''] , use_locking=self._use_locking , ) return tf.no_op() def UpperCamelCase_ ( self : List[Any] , _A : Union[str, Any] , _A : Union[str, Any]=None , **_A : Dict ): _UpperCamelCase , _UpperCamelCase = list(zip(*_A ) ) return super(_A , self ).apply_gradients(zip(_A , _A ) , name=_A , **_A ) def UpperCamelCase_ ( self : Tuple , _A : str , _A : Tuple , _A : str ): if apply_state is None: return self._decayed_lr_t[var_dtype], {} _UpperCamelCase = apply_state or {} _UpperCamelCase = apply_state.get((var_device, var_dtype) ) if coefficients is None: _UpperCamelCase = self._fallback_apply_state(_A , _A ) _UpperCamelCase = coefficients return coefficients["lr_t"], {"apply_state": apply_state} def UpperCamelCase_ ( self : List[str] , _A : Union[str, Any] , _A : List[str] , _A : List[str]=None ): _UpperCamelCase , _UpperCamelCase = self._get_lr(var.device , var.dtype.base_dtype , _A ) _UpperCamelCase = self._decay_weights_op(_A , _A , _A ) with tf.control_dependencies([decay] ): return super(_A , self )._resource_apply_dense(_A , _A , **_A ) def UpperCamelCase_ ( self : Dict , _A : Union[str, Any] , _A : int , _A : List[str] , _A : Optional[Any]=None ): _UpperCamelCase , _UpperCamelCase = self._get_lr(var.device , var.dtype.base_dtype , _A ) _UpperCamelCase = self._decay_weights_op(_A , _A , _A ) with tf.control_dependencies([decay] ): return super(_A , self )._resource_apply_sparse(_A , _A , _A , **_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = super().get_config() config.update({'''weight_decay_rate''': self.weight_decay_rate} ) return config def UpperCamelCase_ ( self : int , _A : Union[str, Any] ): if self.weight_decay_rate == 0: return False if self._include_in_weight_decay: for r in self._include_in_weight_decay: if re.search(_A , _A ) is not None: return True if self._exclude_from_weight_decay: for r in self._exclude_from_weight_decay: if re.search(_A , _A ) is not None: return False return True class lowerCAmelCase_ ( __lowercase ): def __init__( self : Tuple ): _UpperCamelCase = [] _UpperCamelCase = None @property def UpperCamelCase_ ( self : List[Any] ): if self._accum_steps is None: _UpperCamelCase = tf.Variable( tf.constant(0 , dtype=tf.intaa ) , trainable=_A , synchronization=tf.VariableSynchronization.ON_READ , aggregation=tf.VariableAggregation.ONLY_FIRST_REPLICA , ) return self._accum_steps.value() @property def UpperCamelCase_ ( self : int ): if not self._gradients: raise ValueError('''The accumulator should be called first to initialize the gradients''' ) return [gradient.value() if gradient is not None else gradient for gradient in self._gradients] def __call__( self : Optional[int] , _A : Dict ): if not self._gradients: _UpperCamelCase = self.step # Create the step variable. self._gradients.extend( [ tf.Variable( tf.zeros_like(_A ) , trainable=_A , synchronization=tf.VariableSynchronization.ON_READ , aggregation=tf.VariableAggregation.ONLY_FIRST_REPLICA , ) if gradient is not None else gradient for gradient in gradients ] ) if len(_A ) != len(self._gradients ): raise ValueError(F"""Expected {len(self._gradients )} gradients, but got {len(_A )}""" ) for accum_gradient, gradient in zip(self._gradients , _A ): if accum_gradient is not None and gradient is not None: accum_gradient.assign_add(_A ) self._accum_steps.assign_add(1 ) def UpperCamelCase_ ( self : Optional[Any] ): if not self._gradients: return self._accum_steps.assign(0 ) for gradient in self._gradients: if gradient is not None: gradient.assign(tf.zeros_like(_A ) )
71
import torch from diffusers import DPMSolverSDEScheduler from diffusers.utils import torch_device from diffusers.utils.testing_utils import require_torchsde from .test_schedulers import SchedulerCommonTest @require_torchsde class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = (DPMSolverSDEScheduler,) UpperCAmelCase = 10 def UpperCamelCase_ ( self : Tuple , **_A : Union[str, Any] ): _UpperCamelCase = { '''num_train_timesteps''': 1100, '''beta_start''': 0.0001, '''beta_end''': 0.02, '''beta_schedule''': '''linear''', '''noise_sampler_seed''': 0, } config.update(**_A ) return config def UpperCamelCase_ ( self : List[Any] ): for timesteps in [10, 50, 100, 1000]: self.check_over_configs(num_train_timesteps=_A ) def UpperCamelCase_ ( self : List[Any] ): for beta_start, beta_end in zip([0.0_0001, 0.0001, 0.001] , [0.0002, 0.002, 0.02] ): self.check_over_configs(beta_start=_A , beta_end=_A ) def UpperCamelCase_ ( self : List[str] ): for schedule in ["linear", "scaled_linear"]: self.check_over_configs(beta_schedule=_A ) def UpperCamelCase_ ( self : Union[str, Any] ): for prediction_type in ["epsilon", "v_prediction"]: self.check_over_configs(prediction_type=_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for i, t in enumerate(scheduler.timesteps ): _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 167.47_8210_4492_1875 ) < 1e-2 assert abs(result_mean.item() - 0.2178_7059_6456_5277 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 171.59_3521_1181_6406 ) < 1e-2 assert abs(result_mean.item() - 0.2_2342_9068_9229_9652 ) < 1e-3 else: assert abs(result_sum.item() - 162.52_3834_2285_1562 ) < 1e-2 assert abs(result_mean.item() - 0.211_6195_7085_1326 ) < 1e-3 def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config(prediction_type='''v_prediction''' ) _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for i, t in enumerate(scheduler.timesteps ): _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 124.77_1492_0043_9453 ) < 1e-2 assert abs(result_mean.item() - 0.1_6226_2890_1481_6284 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 128.1_6633_6059_5703 ) < 1e-2 assert abs(result_mean.item() - 0.1_6688_3260_0116_7297 ) < 1e-3 else: assert abs(result_sum.item() - 119.8_4875_4882_8125 ) < 1e-2 assert abs(result_mean.item() - 0.1560_5306_6253_6621 ) < 1e-3 def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A ) scheduler.set_timesteps(self.num_inference_steps , device=_A ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter.to(_A ) * scheduler.init_noise_sigma for t in scheduler.timesteps: _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 167.46_9573_9746_0938 ) < 1e-2 assert abs(result_mean.item() - 0.2_1805_9346_0798_2635 ) < 1e-3 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 171.59_3536_3769_5312 ) < 1e-2 assert abs(result_mean.item() - 0.2_2342_9083_8241_5771 ) < 1e-3 else: assert abs(result_sum.item() - 162.52_3834_2285_1562 ) < 1e-2 assert abs(result_mean.item() - 0.211_6195_7085_1326 ) < 1e-3 def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.scheduler_classes[0] _UpperCamelCase = self.get_scheduler_config() _UpperCamelCase = scheduler_class(**_A , use_karras_sigmas=_A ) scheduler.set_timesteps(self.num_inference_steps , device=_A ) _UpperCamelCase = self.dummy_model() _UpperCamelCase = self.dummy_sample_deter.to(_A ) * scheduler.init_noise_sigma _UpperCamelCase = sample.to(_A ) for t in scheduler.timesteps: _UpperCamelCase = scheduler.scale_model_input(_A , _A ) _UpperCamelCase = model(_A , _A ) _UpperCamelCase = scheduler.step(_A , _A , _A ) _UpperCamelCase = output.prev_sample _UpperCamelCase = torch.sum(torch.abs(_A ) ) _UpperCamelCase = torch.mean(torch.abs(_A ) ) if torch_device in ["mps"]: assert abs(result_sum.item() - 176.66_9741_3574_2188 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2 elif torch_device in ["cuda"]: assert abs(result_sum.item() - 177.63_6535_6445_3125 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2 else: assert abs(result_sum.item() - 170.3_1352_2338_8672 ) < 1e-2 assert abs(result_mean.item() - 0.2_3003_8727_3098_1811 ) < 1e-2
71
1
from math import factorial def _snake_case ( __snake_case , __snake_case ): # If either of the conditions are true, the function is being asked # to calculate a factorial of a negative number, which is not possible if n < k or k < 0: raise ValueError('''Please enter positive integers for n and k where n >= k''' ) return factorial(__snake_case ) // (factorial(__snake_case ) * factorial(n - k )) if __name__ == "__main__": print( "The number of five-card hands possible from a standard", f'fifty-two card deck is: {combinations(52, 5)}\n', ) print( "If a class of 40 students must be arranged into groups of", f'4 for group projects, there are {combinations(40, 4)} ways', "to arrange them.\n", ) print( "If 10 teams are competing in a Formula One race, there", f'are {combinations(10, 3)} ways that first, second and', "third place can be awarded.", )
71
import unittest from typing import Tuple import torch from diffusers.utils import floats_tensor, randn_tensor, torch_all_close, torch_device from diffusers.utils.testing_utils import require_torch @require_torch class lowerCAmelCase_ : @property def UpperCamelCase_ ( self : Optional[int] ): return self.get_dummy_input() @property def UpperCamelCase_ ( self : Dict ): if self.block_type == "down": return (4, 32, 16, 16) elif self.block_type == "mid": return (4, 32, 32, 32) elif self.block_type == "up": return (4, 32, 64, 64) raise ValueError(F"""'{self.block_type}' is not a supported block_type. Set it to 'up', 'mid', or 'down'.""" ) def UpperCamelCase_ ( self : Union[str, Any] , _A : List[str]=True , _A : Any=False , _A : Union[str, Any]=False , _A : int=False , ): _UpperCamelCase = 4 _UpperCamelCase = 32 _UpperCamelCase = (32, 32) _UpperCamelCase = torch.manual_seed(0 ) _UpperCamelCase = torch.device(_A ) _UpperCamelCase = (batch_size, num_channels) + sizes _UpperCamelCase = randn_tensor(_A , generator=_A , device=_A ) _UpperCamelCase = {'''hidden_states''': hidden_states} if include_temb: _UpperCamelCase = 128 _UpperCamelCase = randn_tensor((batch_size, temb_channels) , generator=_A , device=_A ) if include_res_hidden_states_tuple: _UpperCamelCase = torch.manual_seed(1 ) _UpperCamelCase = (randn_tensor(_A , generator=_A , device=_A ),) if include_encoder_hidden_states: _UpperCamelCase = floats_tensor((batch_size, 32, 32) ).to(_A ) if include_skip_sample: _UpperCamelCase = randn_tensor(((batch_size, 3) + sizes) , generator=_A , device=_A ) return dummy_input def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = { '''in_channels''': 32, '''out_channels''': 32, '''temb_channels''': 128, } if self.block_type == "up": _UpperCamelCase = 32 if self.block_type == "mid": init_dict.pop('''out_channels''' ) _UpperCamelCase = self.dummy_input return init_dict, inputs_dict def UpperCamelCase_ ( self : Tuple , _A : Union[str, Any] ): _UpperCamelCase , _UpperCamelCase = self.prepare_init_args_and_inputs_for_common() _UpperCamelCase = self.block_class(**_A ) unet_block.to(_A ) unet_block.eval() with torch.no_grad(): _UpperCamelCase = unet_block(**_A ) if isinstance(_A , _A ): _UpperCamelCase = output[0] self.assertEqual(output.shape , self.output_shape ) _UpperCamelCase = output[0, -1, -3:, -3:] _UpperCamelCase = torch.tensor(_A ).to(_A ) assert torch_all_close(output_slice.flatten() , _A , atol=5e-3 ) @unittest.skipIf(torch_device == '''mps''' , '''Training is not supported in mps''' ) def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase , _UpperCamelCase = self.prepare_init_args_and_inputs_for_common() _UpperCamelCase = self.block_class(**_A ) model.to(_A ) model.train() _UpperCamelCase = model(**_A ) if isinstance(_A , _A ): _UpperCamelCase = output[0] _UpperCamelCase = torch.device(_A ) _UpperCamelCase = randn_tensor(output.shape , device=_A ) _UpperCamelCase = torch.nn.functional.mse_loss(_A , _A ) loss.backward()
71
1
from math import cos, sin, sqrt, tau from audio_filters.iir_filter import IIRFilter def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 - _cos) / 2 _UpperCamelCase = 1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 + _cos) / 2 _UpperCamelCase = -1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = _sin / 2 _UpperCamelCase = 0 _UpperCamelCase = -ba _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 1 - alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([ba, ba, ba] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = 1 + alpha * big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha * big_a _UpperCamelCase = 1 + alpha / big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha / big_a _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (pmc + aaa) _UpperCamelCase = 2 * big_a * mpc _UpperCamelCase = big_a * (pmc - aaa) _UpperCamelCase = ppmc + aaa _UpperCamelCase = -2 * pmpc _UpperCamelCase = ppmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (ppmc + aaa) _UpperCamelCase = -2 * big_a * pmpc _UpperCamelCase = big_a * (ppmc - aaa) _UpperCamelCase = pmc + aaa _UpperCamelCase = 2 * mpc _UpperCamelCase = pmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt
71
def _snake_case ( __snake_case ): if not isinstance(__snake_case , __snake_case ): raise TypeError('''Input value must be an \'int\' type''' ) _UpperCamelCase = 0 while number: position += 1 number >>= 1 return position if __name__ == "__main__": import doctest doctest.testmod()
71
1
from __future__ import annotations import unittest from transformers import DebertaVaConfig, is_tf_available from transformers.testing_utils import require_tf, slow from ...test_configuration_common import ConfigTester from ...test_modeling_tf_common import TFModelTesterMixin, ids_tensor, random_attention_mask from ...test_pipeline_mixin import PipelineTesterMixin if is_tf_available(): import tensorflow as tf from transformers import ( TFDebertaVaForMaskedLM, TFDebertaVaForQuestionAnswering, TFDebertaVaForSequenceClassification, TFDebertaVaForTokenClassification, TFDebertaVaModel, ) class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : Optional[Any] , _A : List[str]=13 , _A : Union[str, Any]=7 , _A : int=True , _A : Optional[int]=True , _A : Optional[int]=True , _A : Union[str, Any]=True , _A : Optional[int]=99 , _A : Union[str, Any]=32 , _A : Dict=2 , _A : List[Any]=4 , _A : Optional[Any]=37 , _A : int="gelu" , _A : Optional[int]=0.1 , _A : str=0.1 , _A : List[str]=512 , _A : Optional[Any]=16 , _A : Optional[Any]=2 , _A : Optional[int]=0.02 , _A : str=False , _A : int=True , _A : Any="None" , _A : Dict=3 , _A : List[Any]=4 , _A : Optional[Any]=None , ): _UpperCamelCase = parent _UpperCamelCase = batch_size _UpperCamelCase = seq_length _UpperCamelCase = is_training _UpperCamelCase = use_input_mask _UpperCamelCase = use_token_type_ids _UpperCamelCase = use_labels _UpperCamelCase = vocab_size _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = hidden_dropout_prob _UpperCamelCase = attention_probs_dropout_prob _UpperCamelCase = max_position_embeddings _UpperCamelCase = type_vocab_size _UpperCamelCase = type_sequence_label_size _UpperCamelCase = initializer_range _UpperCamelCase = num_labels _UpperCamelCase = num_choices _UpperCamelCase = relative_attention _UpperCamelCase = position_biased_input _UpperCamelCase = pos_att_type _UpperCamelCase = scope def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size ) _UpperCamelCase = None if self.use_input_mask: _UpperCamelCase = random_attention_mask([self.batch_size, self.seq_length] ) _UpperCamelCase = None if self.use_token_type_ids: _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.type_vocab_size ) _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None if self.use_labels: _UpperCamelCase = ids_tensor([self.batch_size] , self.type_sequence_label_size ) _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.num_labels ) _UpperCamelCase = DebertaVaConfig( vocab_size=self.vocab_size , hidden_size=self.hidden_size , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , relative_attention=self.relative_attention , position_biased_input=self.position_biased_input , initializer_range=self.initializer_range , return_dict=_A , ) return config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels def UpperCamelCase_ ( self : Dict , _A : Tuple , _A : Tuple , _A : Union[str, Any] , _A : List[str] , _A : Optional[int] , _A : int , _A : Optional[Any] ): _UpperCamelCase = TFDebertaVaModel(config=_A ) _UpperCamelCase = {'''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids} _UpperCamelCase = [input_ids, input_mask] _UpperCamelCase = model(_A ) _UpperCamelCase = model(_A ) self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) ) def UpperCamelCase_ ( self : Dict , _A : Optional[int] , _A : Any , _A : Dict , _A : Union[str, Any] , _A : Union[str, Any] , _A : List[Any] , _A : List[str] ): _UpperCamelCase = TFDebertaVaForMaskedLM(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) ) def UpperCamelCase_ ( self : Dict , _A : Dict , _A : List[str] , _A : List[Any] , _A : List[Any] , _A : Optional[Any] , _A : Tuple , _A : int ): _UpperCamelCase = self.num_labels _UpperCamelCase = TFDebertaVaForSequenceClassification(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) ) def UpperCamelCase_ ( self : Tuple , _A : Dict , _A : Optional[int] , _A : Any , _A : List[Any] , _A : Dict , _A : Union[str, Any] , _A : List[str] ): _UpperCamelCase = self.num_labels _UpperCamelCase = TFDebertaVaForTokenClassification(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) ) def UpperCamelCase_ ( self : Dict , _A : Optional[Any] , _A : Optional[int] , _A : Any , _A : List[str] , _A : str , _A : Optional[int] , _A : str ): _UpperCamelCase = TFDebertaVaForQuestionAnswering(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.start_logits.shape , (self.batch_size, self.seq_length) ) self.parent.assertEqual(result.end_logits.shape , (self.batch_size, self.seq_length) ) def UpperCamelCase_ ( self : Any ): _UpperCamelCase = self.prepare_config_and_inputs() ( ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ) = config_and_inputs _UpperCamelCase = {'''input_ids''': input_ids, '''token_type_ids''': token_type_ids, '''attention_mask''': input_mask} return config, inputs_dict @require_tf class lowerCAmelCase_ ( __lowercase, __lowercase, unittest.TestCase ): UpperCAmelCase = ( ( TFDebertaVaModel, TFDebertaVaForMaskedLM, TFDebertaVaForQuestionAnswering, TFDebertaVaForSequenceClassification, TFDebertaVaForTokenClassification, ) if is_tf_available() else () ) UpperCAmelCase = ( { "feature-extraction": TFDebertaVaModel, "fill-mask": TFDebertaVaForMaskedLM, "question-answering": TFDebertaVaForQuestionAnswering, "text-classification": TFDebertaVaForSequenceClassification, "token-classification": TFDebertaVaForTokenClassification, "zero-shot": TFDebertaVaForSequenceClassification, } if is_tf_available() else {} ) UpperCAmelCase = False UpperCAmelCase = False def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = TFDebertaVaModelTester(self ) _UpperCamelCase = ConfigTester(self , config_class=_A , hidden_size=37 ) def UpperCamelCase_ ( self : Any ): self.config_tester.run_common_tests() def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model(*_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_masked_lm(*_A ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_question_answering(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_sequence_classification(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_token_classification(*_A ) @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = TFDebertaVaModel.from_pretrained('''kamalkraj/deberta-v2-xlarge''' ) self.assertIsNotNone(_A ) @require_tf class lowerCAmelCase_ ( unittest.TestCase ): @unittest.skip(reason='''Model not available yet''' ) def UpperCamelCase_ ( self : List[Any] ): pass @slow def UpperCamelCase_ ( self : int ): _UpperCamelCase = TFDebertaVaModel.from_pretrained('''kamalkraj/deberta-v2-xlarge''' ) _UpperCamelCase = tf.constant([[0, 3_1414, 232, 328, 740, 1140, 1_2695, 69, 4_6078, 1588, 2]] ) _UpperCamelCase = tf.constant([[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]] ) _UpperCamelCase = model(_A , attention_mask=_A )[0] _UpperCamelCase = tf.constant( [[[0.2356, 0.1948, 0.0369], [-0.1063, 0.3586, -0.5152], [-0.6399, -0.0259, -0.2525]]] ) tf.debugging.assert_near(output[:, 1:4, 1:4] , _A , atol=1e-4 )
71
import argparse import json import os import fairseq import torch from torch import nn from transformers import ( SpeechaTextaConfig, SpeechaTextaForCausalLM, SpeechaTextaTokenizer, SpeechEncoderDecoderConfig, SpeechEncoderDecoderModel, WavaVecaConfig, WavaVecaFeatureExtractor, WavaVecaModel, logging, ) logging.set_verbosity_info() _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "post_extract_proj": "feature_projection.projection", "encoder.pos_conv.0": "encoder.pos_conv_embed.conv", "self_attn.k_proj": "encoder.layers.*.attention.k_proj", "self_attn.v_proj": "encoder.layers.*.attention.v_proj", "self_attn.q_proj": "encoder.layers.*.attention.q_proj", "self_attn.out_proj": "encoder.layers.*.attention.out_proj", "self_attn_layer_norm": "encoder.layers.*.layer_norm", "fc1": "encoder.layers.*.feed_forward.intermediate_dense", "fc2": "encoder.layers.*.feed_forward.output_dense", "final_layer_norm": "encoder.layers.*.final_layer_norm", "encoder.layer_norm": "encoder.layer_norm", "w2v_model.layer_norm": "feature_projection.layer_norm", "quantizer.weight_proj": "quantizer.weight_proj", "quantizer.vars": "quantizer.codevectors", "project_q": "project_q", "final_proj": "project_hid", "w2v_encoder.proj": "lm_head", "mask_emb": "masked_spec_embed", } _lowerCAmelCase = [ "lm_head", "quantizer.weight_proj", "quantizer.codevectors", "project_q", "project_hid", ] def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): for attribute in key.split('''.''' ): _UpperCamelCase = getattr(__snake_case , __snake_case ) if weight_type is not None: _UpperCamelCase = getattr(__snake_case , __snake_case ).shape else: _UpperCamelCase = hf_pointer.shape assert hf_shape == value.shape, ( f"""Shape of hf {key + "." + weight_type if weight_type is not None else ""} is {hf_shape}, but should be""" f""" {value.shape} for {full_name}""" ) if weight_type == "weight": _UpperCamelCase = value elif weight_type == "weight_g": _UpperCamelCase = value elif weight_type == "weight_v": _UpperCamelCase = value elif weight_type == "bias": _UpperCamelCase = value else: _UpperCamelCase = value logger.info(f"""{key + "." + weight_type if weight_type is not None else ""} was initialized from {full_name}.""" ) def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = [] _UpperCamelCase = fairseq_model.state_dict() _UpperCamelCase = hf_model.feature_extractor # if encoder has different dim to decoder -> use proj_weight _UpperCamelCase = None for name, value in fairseq_dict.items(): _UpperCamelCase = False if "conv_layers" in name: load_conv_layer( __snake_case , __snake_case , __snake_case , __snake_case , hf_model.config.feat_extract_norm == '''group''' , ) _UpperCamelCase = True elif name.split('''.''' )[0] == "proj": _UpperCamelCase = fairseq_model.proj _UpperCamelCase = True else: for key, mapped_key in MAPPING.items(): if key in name or key.split('''w2v_model.''' )[-1] == name.split('''.''' )[0]: _UpperCamelCase = True if "*" in mapped_key: _UpperCamelCase = name.split(__snake_case )[0].split('''.''' )[-2] _UpperCamelCase = mapped_key.replace('''*''' , __snake_case ) if "weight_g" in name: _UpperCamelCase = '''weight_g''' elif "weight_v" in name: _UpperCamelCase = '''weight_v''' elif "bias" in name: _UpperCamelCase = '''bias''' elif "weight" in name: _UpperCamelCase = '''weight''' else: _UpperCamelCase = None set_recursively(__snake_case , __snake_case , __snake_case , __snake_case , __snake_case ) continue if not is_used: unused_weights.append(__snake_case ) logger.warning(f"""Unused weights: {unused_weights}""" ) return proj_weight def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = full_name.split('''conv_layers.''' )[-1] _UpperCamelCase = name.split('''.''' ) _UpperCamelCase = int(items[0] ) _UpperCamelCase = int(items[1] ) if type_id == 0: if "bias" in name: assert value.shape == feature_extractor.conv_layers[layer_id].conv.bias.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor.conv_layers[layer_id].conv.bias.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract conv layer {layer_id} was initialized from {full_name}.""" ) elif "weight" in name: assert value.shape == feature_extractor.conv_layers[layer_id].conv.weight.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor.conv_layers[layer_id].conv.weight.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract conv layer {layer_id} was initialized from {full_name}.""" ) elif (type_id == 2 and not use_group_norm) or (type_id == 2 and layer_id == 0 and use_group_norm): if "bias" in name: assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.bias.data.shape, ( f"""{full_name} has size {value.shape}, but {feature_extractor[layer_id].layer_norm.bias.data.shape} was""" " found." ) _UpperCamelCase = value logger.info(f"""Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.""" ) elif "weight" in name: assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.weight.data.shape, ( f"""{full_name} has size {value.shape}, but""" f""" {feature_extractor[layer_id].layer_norm.weight.data.shape} was found.""" ) _UpperCamelCase = value logger.info(f"""Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.""" ) else: unused_weights.append(__snake_case ) def _snake_case ( __snake_case ): _UpperCamelCase , _UpperCamelCase = emb.weight.shape _UpperCamelCase = nn.Linear(__snake_case , __snake_case , bias=__snake_case ) _UpperCamelCase = emb.weight.data return lin_layer def _snake_case ( __snake_case ): with open(__snake_case , '''r''' , encoding='''utf-8''' ) as f: _UpperCamelCase = f.readlines() _UpperCamelCase = [line.split(''' ''' )[0] for line in lines] _UpperCamelCase = len(__snake_case ) _UpperCamelCase = { '''<s>''': 0, '''<pad>''': 1, '''</s>''': 2, '''<unk>''': 3, } vocab_dict.update(dict(zip(__snake_case , range(4 , num_words + 4 ) ) ) ) return vocab_dict @torch.no_grad() def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , ): _UpperCamelCase = WavaVecaConfig.from_pretrained(__snake_case ) _UpperCamelCase = SpeechaTextaConfig.from_pretrained( __snake_case , vocab_size=__snake_case , decoder_layers=__snake_case , do_stable_layer_norm=__snake_case ) _UpperCamelCase = WavaVecaFeatureExtractor( feature_size=1 , sampling_rate=16000 , padding_value=0 , do_normalize=__snake_case , return_attention_mask=__snake_case , ) _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = fairseq.checkpoint_utils.load_model_ensemble_and_task( [checkpoint_path] , arg_overrides={'''data''': '''/'''.join(dict_path.split('''/''' )[:-1] )} ) _UpperCamelCase = model[0].eval() # set weights for wav2vec2 encoder _UpperCamelCase = WavaVecaModel(__snake_case ) _UpperCamelCase = recursively_load_weights_wavaveca(model.encoder , __snake_case ) _UpperCamelCase = SpeechaTextaForCausalLM(__snake_case ) _UpperCamelCase , _UpperCamelCase = hf_decoder.model.decoder.load_state_dict(model.decoder.state_dict() , strict=__snake_case ) # set output linear layer unexpected_keys.remove('''embed_out''' ) _UpperCamelCase = nn.Parameter(model.decoder.embed_out.detach() ) # layer norm is init to identity matrix so leaving it is fine logger.warning(f"""The following keys are missing when loading the decoder weights: {missing_keys}""" ) logger.warning(f"""The following keys are unexpected when loading the decoder weights: {unexpected_keys}""" ) _UpperCamelCase = SpeechEncoderDecoderModel(encoder=__snake_case , decoder=__snake_case ) _UpperCamelCase = False # add projection layer _UpperCamelCase = nn.Parameter(projection_layer.weight ) _UpperCamelCase = nn.Parameter(projection_layer.bias ) _UpperCamelCase = create_vocab_dict(__snake_case ) with open(os.path.join(__snake_case , '''vocab.json''' ) , '''w''' ) as fp: json.dump(__snake_case , __snake_case ) _UpperCamelCase = SpeechaTextaTokenizer(os.path.join(__snake_case , '''vocab.json''' ) ) tokenizer.save_pretrained(__snake_case ) _UpperCamelCase = hf_wavavec.config.to_dict() _UpperCamelCase = tokenizer.pad_token_id _UpperCamelCase = tokenizer.bos_token_id _UpperCamelCase = tokenizer.eos_token_id _UpperCamelCase = '''speech_to_text_2''' _UpperCamelCase = '''wav2vec2''' _UpperCamelCase = SpeechEncoderDecoderConfig.from_dict(__snake_case ) hf_wavavec.save_pretrained(__snake_case ) feature_extractor.save_pretrained(__snake_case ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() parser.add_argument("--pytorch_dump_folder_path", default=None, type=str, help="Path to the output PyTorch model.") parser.add_argument("--checkpoint_path", default=None, type=str, help="Path to fairseq checkpoint") parser.add_argument("--dict_path", default=None, type=str, help="Path to dict of fine-tuned model") parser.add_argument( "--encoder_config_path", default="facebook/wav2vec2-large-lv60", type=str, help="Path to hf encoder wav2vec2 checkpoint config", ) parser.add_argument( "--decoder_config_path", default="facebook/s2t-small-mustc-en-fr-st", type=str, help="Path to hf decoder s2t checkpoint config", ) parser.add_argument("--vocab_size", default=10_224, type=int, help="Vocab size of decoder") parser.add_argument("--num_decoder_layers", default=7, type=int, help="Number of decoder layers") _lowerCAmelCase = parser.parse_args() convert_wavaveca_checkpoint( args.checkpoint_path, args.pytorch_dump_folder_path, args.dict_path, encoder_config_path=args.encoder_config_path, decoder_config_path=args.decoder_config_path, vocab_size=args.vocab_size, num_decoder_layers=args.num_decoder_layers, )
71
1
def _snake_case ( __snake_case ): _UpperCamelCase = 0 while num > 0: digit_sum += num % 10 num //= 10 return digit_sum def _snake_case ( __snake_case = 100 ): _UpperCamelCase = 1 _UpperCamelCase = 2 for i in range(2 , max_n + 1 ): _UpperCamelCase = pre_numerator _UpperCamelCase = 2 * i // 3 if i % 3 == 0 else 1 _UpperCamelCase = cur_numerator _UpperCamelCase = e_cont * pre_numerator + temp return sum_digits(__snake_case ) if __name__ == "__main__": print(f'{solution() = }')
71
from __future__ import annotations import unittest from transformers import DebertaVaConfig, is_tf_available from transformers.testing_utils import require_tf, slow from ...test_configuration_common import ConfigTester from ...test_modeling_tf_common import TFModelTesterMixin, ids_tensor, random_attention_mask from ...test_pipeline_mixin import PipelineTesterMixin if is_tf_available(): import tensorflow as tf from transformers import ( TFDebertaVaForMaskedLM, TFDebertaVaForQuestionAnswering, TFDebertaVaForSequenceClassification, TFDebertaVaForTokenClassification, TFDebertaVaModel, ) class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : Optional[Any] , _A : List[str]=13 , _A : Union[str, Any]=7 , _A : int=True , _A : Optional[int]=True , _A : Optional[int]=True , _A : Union[str, Any]=True , _A : Optional[int]=99 , _A : Union[str, Any]=32 , _A : Dict=2 , _A : List[Any]=4 , _A : Optional[Any]=37 , _A : int="gelu" , _A : Optional[int]=0.1 , _A : str=0.1 , _A : List[str]=512 , _A : Optional[Any]=16 , _A : Optional[Any]=2 , _A : Optional[int]=0.02 , _A : str=False , _A : int=True , _A : Any="None" , _A : Dict=3 , _A : List[Any]=4 , _A : Optional[Any]=None , ): _UpperCamelCase = parent _UpperCamelCase = batch_size _UpperCamelCase = seq_length _UpperCamelCase = is_training _UpperCamelCase = use_input_mask _UpperCamelCase = use_token_type_ids _UpperCamelCase = use_labels _UpperCamelCase = vocab_size _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = hidden_dropout_prob _UpperCamelCase = attention_probs_dropout_prob _UpperCamelCase = max_position_embeddings _UpperCamelCase = type_vocab_size _UpperCamelCase = type_sequence_label_size _UpperCamelCase = initializer_range _UpperCamelCase = num_labels _UpperCamelCase = num_choices _UpperCamelCase = relative_attention _UpperCamelCase = position_biased_input _UpperCamelCase = pos_att_type _UpperCamelCase = scope def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size ) _UpperCamelCase = None if self.use_input_mask: _UpperCamelCase = random_attention_mask([self.batch_size, self.seq_length] ) _UpperCamelCase = None if self.use_token_type_ids: _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.type_vocab_size ) _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None if self.use_labels: _UpperCamelCase = ids_tensor([self.batch_size] , self.type_sequence_label_size ) _UpperCamelCase = ids_tensor([self.batch_size, self.seq_length] , self.num_labels ) _UpperCamelCase = DebertaVaConfig( vocab_size=self.vocab_size , hidden_size=self.hidden_size , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , relative_attention=self.relative_attention , position_biased_input=self.position_biased_input , initializer_range=self.initializer_range , return_dict=_A , ) return config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels def UpperCamelCase_ ( self : Dict , _A : Tuple , _A : Tuple , _A : Union[str, Any] , _A : List[str] , _A : Optional[int] , _A : int , _A : Optional[Any] ): _UpperCamelCase = TFDebertaVaModel(config=_A ) _UpperCamelCase = {'''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids} _UpperCamelCase = [input_ids, input_mask] _UpperCamelCase = model(_A ) _UpperCamelCase = model(_A ) self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) ) def UpperCamelCase_ ( self : Dict , _A : Optional[int] , _A : Any , _A : Dict , _A : Union[str, Any] , _A : Union[str, Any] , _A : List[Any] , _A : List[str] ): _UpperCamelCase = TFDebertaVaForMaskedLM(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) ) def UpperCamelCase_ ( self : Dict , _A : Dict , _A : List[str] , _A : List[Any] , _A : List[Any] , _A : Optional[Any] , _A : Tuple , _A : int ): _UpperCamelCase = self.num_labels _UpperCamelCase = TFDebertaVaForSequenceClassification(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) ) def UpperCamelCase_ ( self : Tuple , _A : Dict , _A : Optional[int] , _A : Any , _A : List[Any] , _A : Dict , _A : Union[str, Any] , _A : List[str] ): _UpperCamelCase = self.num_labels _UpperCamelCase = TFDebertaVaForTokenClassification(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) ) def UpperCamelCase_ ( self : Dict , _A : Optional[Any] , _A : Optional[int] , _A : Any , _A : List[str] , _A : str , _A : Optional[int] , _A : str ): _UpperCamelCase = TFDebertaVaForQuestionAnswering(config=_A ) _UpperCamelCase = { '''input_ids''': input_ids, '''attention_mask''': input_mask, '''token_type_ids''': token_type_ids, } _UpperCamelCase = model(_A ) self.parent.assertEqual(result.start_logits.shape , (self.batch_size, self.seq_length) ) self.parent.assertEqual(result.end_logits.shape , (self.batch_size, self.seq_length) ) def UpperCamelCase_ ( self : Any ): _UpperCamelCase = self.prepare_config_and_inputs() ( ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ( _UpperCamelCase ) , ) = config_and_inputs _UpperCamelCase = {'''input_ids''': input_ids, '''token_type_ids''': token_type_ids, '''attention_mask''': input_mask} return config, inputs_dict @require_tf class lowerCAmelCase_ ( __lowercase, __lowercase, unittest.TestCase ): UpperCAmelCase = ( ( TFDebertaVaModel, TFDebertaVaForMaskedLM, TFDebertaVaForQuestionAnswering, TFDebertaVaForSequenceClassification, TFDebertaVaForTokenClassification, ) if is_tf_available() else () ) UpperCAmelCase = ( { "feature-extraction": TFDebertaVaModel, "fill-mask": TFDebertaVaForMaskedLM, "question-answering": TFDebertaVaForQuestionAnswering, "text-classification": TFDebertaVaForSequenceClassification, "token-classification": TFDebertaVaForTokenClassification, "zero-shot": TFDebertaVaForSequenceClassification, } if is_tf_available() else {} ) UpperCAmelCase = False UpperCAmelCase = False def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = TFDebertaVaModelTester(self ) _UpperCamelCase = ConfigTester(self , config_class=_A , hidden_size=37 ) def UpperCamelCase_ ( self : Any ): self.config_tester.run_common_tests() def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model(*_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_masked_lm(*_A ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_question_answering(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_sequence_classification(*_A ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_token_classification(*_A ) @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = TFDebertaVaModel.from_pretrained('''kamalkraj/deberta-v2-xlarge''' ) self.assertIsNotNone(_A ) @require_tf class lowerCAmelCase_ ( unittest.TestCase ): @unittest.skip(reason='''Model not available yet''' ) def UpperCamelCase_ ( self : List[Any] ): pass @slow def UpperCamelCase_ ( self : int ): _UpperCamelCase = TFDebertaVaModel.from_pretrained('''kamalkraj/deberta-v2-xlarge''' ) _UpperCamelCase = tf.constant([[0, 3_1414, 232, 328, 740, 1140, 1_2695, 69, 4_6078, 1588, 2]] ) _UpperCamelCase = tf.constant([[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]] ) _UpperCamelCase = model(_A , attention_mask=_A )[0] _UpperCamelCase = tf.constant( [[[0.2356, 0.1948, 0.0369], [-0.1063, 0.3586, -0.5152], [-0.6399, -0.0259, -0.2525]]] ) tf.debugging.assert_near(output[:, 1:4, 1:4] , _A , atol=1e-4 )
71
1
import argparse from collections import defaultdict import yaml _lowerCAmelCase = "docs/source/en/_toctree.yml" def _snake_case ( __snake_case ): _UpperCamelCase = defaultdict(__snake_case ) for doc in model_doc: counts[doc["local"]] += 1 _UpperCamelCase = [key for key, value in counts.items() if value > 1] _UpperCamelCase = [] for duplicate_key in duplicates: _UpperCamelCase = list({doc['''title'''] for doc in model_doc if doc['''local'''] == duplicate_key} ) if len(__snake_case ) > 1: raise ValueError( f"""{duplicate_key} is present several times in the documentation table of content at """ '''`docs/source/en/_toctree.yml` with different *Title* values. Choose one of those and remove the ''' '''others.''' ) # Only add this once new_doc.append({'''local''': duplicate_key, '''title''': titles[0]} ) # Add none duplicate-keys new_doc.extend([doc for doc in model_doc if counts[doc['''local''']] == 1] ) # Sort return sorted(__snake_case , key=lambda __snake_case : s["title"].lower() ) def _snake_case ( __snake_case=False ): with open(__snake_case , encoding='''utf-8''' ) as f: _UpperCamelCase = yaml.safe_load(f.read() ) # Get to the API doc _UpperCamelCase = 0 while content[api_idx]["title"] != "API": api_idx += 1 _UpperCamelCase = content[api_idx]['''sections'''] # Then to the model doc _UpperCamelCase = 0 while api_doc[model_idx]["title"] != "Models": model_idx += 1 _UpperCamelCase = api_doc[model_idx]['''sections'''] _UpperCamelCase = [(idx, section) for idx, section in enumerate(__snake_case ) if '''sections''' in section] _UpperCamelCase = False for idx, modality_doc in modalities_docs: _UpperCamelCase = modality_doc['''sections'''] _UpperCamelCase = clean_model_doc_toc(__snake_case ) if old_modality_doc != new_modality_doc: _UpperCamelCase = True if overwrite: _UpperCamelCase = new_modality_doc if diff: if overwrite: _UpperCamelCase = model_doc _UpperCamelCase = api_doc with open(__snake_case , '''w''' , encoding='''utf-8''' ) as f: f.write(yaml.dump(__snake_case , allow_unicode=__snake_case ) ) else: raise ValueError( '''The model doc part of the table of content is not properly sorted, run `make style` to fix this.''' ) if __name__ == "__main__": _lowerCAmelCase = argparse.ArgumentParser() parser.add_argument("--fix_and_overwrite", action="store_true", help="Whether to fix inconsistencies.") _lowerCAmelCase = parser.parse_args() check_model_doc(args.fix_and_overwrite)
71
def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): # Return True if there is node that has not iterated. _UpperCamelCase = [False] * len(__snake_case ) _UpperCamelCase = [] queue.append(__snake_case ) _UpperCamelCase = True while queue: _UpperCamelCase = queue.pop(0 ) for ind in range(len(graph[u] ) ): if visited[ind] is False and graph[u][ind] > 0: queue.append(__snake_case ) _UpperCamelCase = True _UpperCamelCase = u return visited[t] def _snake_case ( __snake_case , __snake_case , __snake_case ): # This array is filled by BFS and to store path _UpperCamelCase = [-1] * (len(__snake_case )) _UpperCamelCase = 0 while bfs(__snake_case , __snake_case , __snake_case , __snake_case ): _UpperCamelCase = float('''Inf''' ) _UpperCamelCase = sink while s != source: # Find the minimum value in select path _UpperCamelCase = min(__snake_case , graph[parent[s]][s] ) _UpperCamelCase = parent[s] max_flow += path_flow _UpperCamelCase = sink while v != source: _UpperCamelCase = parent[v] graph[u][v] -= path_flow graph[v][u] += path_flow _UpperCamelCase = parent[v] return max_flow _lowerCAmelCase = [ [0, 16, 13, 0, 0, 0], [0, 0, 10, 12, 0, 0], [0, 4, 0, 0, 14, 0], [0, 0, 9, 0, 0, 20], [0, 0, 0, 7, 0, 4], [0, 0, 0, 0, 0, 0], ] _lowerCAmelCase, _lowerCAmelCase = 0, 5 print(ford_fulkerson(graph, source, sink))
71
1
import copy from typing import Any, Dict, List, Optional, Union import numpy as np import torch from ...audio_utils import mel_filter_bank, spectrogram, window_function from ...feature_extraction_sequence_utils import SequenceFeatureExtractor from ...feature_extraction_utils import BatchFeature from ...utils import TensorType, logging _lowerCAmelCase = logging.get_logger(__name__) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = ["input_features", "is_longer"] def __init__( self : Tuple , _A : Tuple=64 , _A : Any=4_8000 , _A : List[str]=480 , _A : Optional[int]=10 , _A : List[Any]=1024 , _A : str=0.0 , _A : Union[str, Any]=False , _A : float = 0 , _A : float = 1_4000 , _A : int = None , _A : str = "fusion" , _A : str = "repeatpad" , **_A : Union[str, Any] , ): super().__init__( feature_size=_A , sampling_rate=_A , padding_value=_A , return_attention_mask=_A , **_A , ) _UpperCamelCase = top_db _UpperCamelCase = truncation _UpperCamelCase = padding _UpperCamelCase = fft_window_size _UpperCamelCase = (fft_window_size >> 1) + 1 _UpperCamelCase = hop_length _UpperCamelCase = max_length_s _UpperCamelCase = max_length_s * sampling_rate _UpperCamelCase = sampling_rate _UpperCamelCase = frequency_min _UpperCamelCase = frequency_max _UpperCamelCase = mel_filter_bank( num_frequency_bins=self.nb_frequency_bins , num_mel_filters=_A , min_frequency=_A , max_frequency=_A , sampling_rate=_A , norm=_A , mel_scale='''htk''' , ) _UpperCamelCase = mel_filter_bank( num_frequency_bins=self.nb_frequency_bins , num_mel_filters=_A , min_frequency=_A , max_frequency=_A , sampling_rate=_A , norm='''slaney''' , mel_scale='''slaney''' , ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = copy.deepcopy(self.__dict__ ) _UpperCamelCase = self.__class__.__name__ if "mel_filters" in output: del output["mel_filters"] if "mel_filters_slaney" in output: del output["mel_filters_slaney"] return output def UpperCamelCase_ ( self : List[Any] , _A : np.array , _A : Optional[np.array] = None ): _UpperCamelCase = spectrogram( _A , window_function(self.fft_window_size , '''hann''' ) , frame_length=self.fft_window_size , hop_length=self.hop_length , power=2.0 , mel_filters=_A , log_mel='''dB''' , ) return log_mel_spectrogram.T def UpperCamelCase_ ( self : Union[str, Any] , _A : List[Any] , _A : Optional[Any] , _A : List[Any] ): _UpperCamelCase = np.array_split(list(range(0 , total_frames - chunk_frames + 1 ) ) , 3 ) if len(ranges[1] ) == 0: # if the audio is too short, we just use the first chunk _UpperCamelCase = [0] if len(ranges[2] ) == 0: # if the audio is too short, we just use the first chunk _UpperCamelCase = [0] # randomly choose index for each part _UpperCamelCase = np.random.choice(ranges[0] ) _UpperCamelCase = np.random.choice(ranges[1] ) _UpperCamelCase = np.random.choice(ranges[2] ) _UpperCamelCase = mel[idx_front : idx_front + chunk_frames, :] _UpperCamelCase = mel[idx_middle : idx_middle + chunk_frames, :] _UpperCamelCase = mel[idx_back : idx_back + chunk_frames, :] _UpperCamelCase = torch.tensor(mel[None, None, :] ) _UpperCamelCase = torch.nn.functional.interpolate( _A , size=[chunk_frames, 64] , mode='''bilinear''' , align_corners=_A ) _UpperCamelCase = mel_shrink[0][0].numpy() _UpperCamelCase = np.stack([mel_shrink, mel_chunk_front, mel_chunk_middle, mel_chunk_back] , axis=0 ) return mel_fusion def UpperCamelCase_ ( self : Optional[int] , _A : np.array , _A : int , _A : str , _A : int ): if waveform.shape[0] > max_length: if truncation == "rand_trunc": _UpperCamelCase = True # random crop to max_length (for compatibility) -> this should be handled by self.pad _UpperCamelCase = len(_A ) - max_length _UpperCamelCase = np.random.randint(0 , overflow + 1 ) _UpperCamelCase = waveform[idx : idx + max_length] _UpperCamelCase = self._np_extract_fbank_features(_A , self.mel_filters_slaney )[None, :] elif truncation == "fusion": _UpperCamelCase = self._np_extract_fbank_features(_A , self.mel_filters ) _UpperCamelCase = max_length // self.hop_length + 1 # the +1 related to how the spectrogram is computed _UpperCamelCase = mel.shape[0] if chunk_frames == total_frames: # there is a corner case where the audio length is larger than max_length but smaller than max_length+hop_length. # In this case, we just use the whole audio. _UpperCamelCase = np.stack([mel, mel, mel, mel] , axis=0 ) _UpperCamelCase = False else: _UpperCamelCase = self._random_mel_fusion(_A , _A , _A ) _UpperCamelCase = True else: raise NotImplementedError(F"""data_truncating {truncation} not implemented""" ) else: _UpperCamelCase = False # only use repeat as a new possible value for padding. you repeat the audio before applying the usual max_length padding if waveform.shape[0] < max_length: if padding == "repeat": _UpperCamelCase = int(max_length / len(_A ) ) _UpperCamelCase = np.stack(np.tile(_A , n_repeat + 1 ) )[:max_length] if padding == "repeatpad": _UpperCamelCase = int(max_length / len(_A ) ) _UpperCamelCase = np.stack(np.tile(_A , _A ) ) _UpperCamelCase = np.pad(_A , (0, max_length - waveform.shape[0]) , mode='''constant''' , constant_values=0 ) if truncation == "fusion": _UpperCamelCase = self._np_extract_fbank_features(_A , self.mel_filters ) _UpperCamelCase = np.stack([input_mel, input_mel, input_mel, input_mel] , axis=0 ) else: _UpperCamelCase = self._np_extract_fbank_features(_A , self.mel_filters_slaney )[None, :] return input_mel, longer def __call__( self : str , _A : Union[np.ndarray, List[float], List[np.ndarray], List[List[float]]] , _A : str = None , _A : Optional[str] = None , _A : Optional[int] = None , _A : Optional[int] = None , _A : Optional[Union[str, TensorType]] = None , **_A : str , ): _UpperCamelCase = truncation if truncation is not None else self.truncation _UpperCamelCase = padding if padding else self.padding if sampling_rate is not None: if sampling_rate != self.sampling_rate: raise ValueError( F"""The model corresponding to this feature extractor: {self.__class__.__name__} was trained using a""" F""" sampling rate of {self.sampling_rate}. Please make sure that the provided `raw_speech` input""" F""" was sampled with {self.sampling_rate} and not {sampling_rate}.""" ) else: logger.warning( '''It is strongly recommended to pass the `sampling_rate` argument to this function. ''' '''Failing to do so can result in silent errors that might be hard to debug.''' ) _UpperCamelCase = isinstance(_A , np.ndarray ) and len(raw_speech.shape ) > 1 if is_batched_numpy and len(raw_speech.shape ) > 2: raise ValueError(F"""Only mono-channel audio is supported for input to {self}""" ) _UpperCamelCase = is_batched_numpy or ( isinstance(_A , (list, tuple) ) and (isinstance(raw_speech[0] , (np.ndarray, tuple, list) )) ) if is_batched: _UpperCamelCase = [np.asarray(_A , dtype=np.floataa ) for speech in raw_speech] elif not is_batched and not isinstance(_A , np.ndarray ): _UpperCamelCase = np.asarray(_A , dtype=np.floataa ) elif isinstance(_A , np.ndarray ) and raw_speech.dtype is np.dtype(np.floataa ): _UpperCamelCase = raw_speech.astype(np.floataa ) # always return batch if not is_batched: _UpperCamelCase = [np.asarray(_A )] # convert to mel spectrogram, truncate and pad if needed. _UpperCamelCase = [ self._get_input_mel(_A , max_length if max_length else self.nb_max_samples , _A , _A ) for waveform in raw_speech ] _UpperCamelCase = [] _UpperCamelCase = [] for mel, longer in padded_inputs: input_mel.append(_A ) is_longer.append(_A ) if truncation == "fusion" and sum(_A ) == 0: # if no audio is longer than 10s, then randomly select one audio to be longer _UpperCamelCase = np.random.randint(0 , len(_A ) ) _UpperCamelCase = True if isinstance(input_mel[0] , _A ): _UpperCamelCase = [np.asarray(_A , dtype=np.floataa ) for feature in input_mel] # is_longer is a list of bool _UpperCamelCase = [[longer] for longer in is_longer] _UpperCamelCase = {'''input_features''': input_mel, '''is_longer''': is_longer} _UpperCamelCase = BatchFeature(_A ) if return_tensors is not None: _UpperCamelCase = input_features.convert_to_tensors(_A ) return input_features
71
from typing import TYPE_CHECKING from ...utils import ( OptionalDependencyNotAvailable, _LazyModule, is_flax_available, is_tf_available, is_torch_available, ) _lowerCAmelCase = {"configuration_unispeech": ["UNISPEECH_PRETRAINED_CONFIG_ARCHIVE_MAP", "UniSpeechConfig"]} try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "UNISPEECH_PRETRAINED_MODEL_ARCHIVE_LIST", "UniSpeechForCTC", "UniSpeechForPreTraining", "UniSpeechForSequenceClassification", "UniSpeechModel", "UniSpeechPreTrainedModel", ] if TYPE_CHECKING: from .configuration_unispeech import UNISPEECH_PRETRAINED_CONFIG_ARCHIVE_MAP, UniSpeechConfig try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_unispeech import ( UNISPEECH_PRETRAINED_MODEL_ARCHIVE_LIST, UniSpeechForCTC, UniSpeechForPreTraining, UniSpeechForSequenceClassification, UniSpeechModel, UniSpeechPreTrainedModel, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
1
from typing import Any def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , ): _validation( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , ) # Creates data structures and fill initial step _UpperCamelCase = {} _UpperCamelCase = {} for state in states_space: _UpperCamelCase = observations_space[0] _UpperCamelCase = ( initial_probabilities[state] * emission_probabilities[state][observation] ) _UpperCamelCase = None # Fills the data structure with the probabilities of # different transitions and pointers to previous states for o in range(1 , len(__snake_case ) ): _UpperCamelCase = observations_space[o] _UpperCamelCase = observations_space[o - 1] for state in states_space: # Calculates the argmax for probability function _UpperCamelCase = '''''' _UpperCamelCase = -1 for k_state in states_space: _UpperCamelCase = ( probabilities[(k_state, prior_observation)] * transition_probabilities[k_state][state] * emission_probabilities[state][observation] ) if probability > max_probability: _UpperCamelCase = probability _UpperCamelCase = k_state # Update probabilities and pointers dicts _UpperCamelCase = ( probabilities[(arg_max, prior_observation)] * transition_probabilities[arg_max][state] * emission_probabilities[state][observation] ) _UpperCamelCase = arg_max # The final observation _UpperCamelCase = observations_space[len(__snake_case ) - 1] # argmax for given final observation _UpperCamelCase = '''''' _UpperCamelCase = -1 for k_state in states_space: _UpperCamelCase = probabilities[(k_state, final_observation)] if probability > max_probability: _UpperCamelCase = probability _UpperCamelCase = k_state _UpperCamelCase = arg_max # Process pointers backwards _UpperCamelCase = last_state _UpperCamelCase = [] for o in range(len(__snake_case ) - 1 , -1 , -1 ): result.append(__snake_case ) _UpperCamelCase = pointers[previous, observations_space[o]] result.reverse() return result def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , ): _validate_not_empty( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , ) _validate_lists(__snake_case , __snake_case ) _validate_dicts( __snake_case , __snake_case , __snake_case ) def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , ): if not all( [ observations_space, states_space, initial_probabilities, transition_probabilities, emission_probabilities, ] ): raise ValueError('''There\'s an empty parameter''' ) def _snake_case ( __snake_case , __snake_case ): _validate_list(__snake_case , '''observations_space''' ) _validate_list(__snake_case , '''states_space''' ) def _snake_case ( __snake_case , __snake_case ): if not isinstance(_object , __snake_case ): _UpperCamelCase = f"""{var_name} must be a list""" raise ValueError(__snake_case ) else: for x in _object: if not isinstance(__snake_case , __snake_case ): _UpperCamelCase = f"""{var_name} must be a list of strings""" raise ValueError(__snake_case ) def _snake_case ( __snake_case , __snake_case , __snake_case , ): _validate_dict(__snake_case , '''initial_probabilities''' , __snake_case ) _validate_nested_dict(__snake_case , '''transition_probabilities''' ) _validate_nested_dict(__snake_case , '''emission_probabilities''' ) def _snake_case ( __snake_case , __snake_case ): _validate_dict(_object , __snake_case , __snake_case ) for x in _object.values(): _validate_dict(__snake_case , __snake_case , __snake_case , __snake_case ) def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = False ): if not isinstance(_object , __snake_case ): _UpperCamelCase = f"""{var_name} must be a dict""" raise ValueError(__snake_case ) if not all(isinstance(__snake_case , __snake_case ) for x in _object ): _UpperCamelCase = f"""{var_name} all keys must be strings""" raise ValueError(__snake_case ) if not all(isinstance(__snake_case , __snake_case ) for x in _object.values() ): _UpperCamelCase = '''nested dictionary ''' if nested else '''''' _UpperCamelCase = f"""{var_name} {nested_text}all values must be {value_type.__name__}""" raise ValueError(__snake_case ) if __name__ == "__main__": from doctest import testmod testmod()
71
import json import os import shutil import tempfile import unittest import numpy as np from transformers import BertTokenizerFast from transformers.models.bert.tokenization_bert import VOCAB_FILES_NAMES, BertTokenizer from transformers.testing_utils import require_tokenizers, require_vision from transformers.utils import IMAGE_PROCESSOR_NAME, is_vision_available if is_vision_available(): from PIL import Image from transformers import VisionTextDualEncoderProcessor, ViTImageProcessor @require_tokenizers @require_vision class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = tempfile.mkdtemp() # fmt: off _UpperCamelCase = ['''[UNK]''', '''[CLS]''', '''[SEP]''', '''[PAD]''', '''[MASK]''', '''want''', '''##want''', '''##ed''', '''wa''', '''un''', '''runn''', '''##ing''', ''',''', '''low''', '''lowest'''] # fmt: on _UpperCamelCase = os.path.join(self.tmpdirname , VOCAB_FILES_NAMES['''vocab_file'''] ) with open(self.vocab_file , '''w''' , encoding='''utf-8''' ) as vocab_writer: vocab_writer.write(''''''.join([x + '''\n''' for x in vocab_tokens] ) ) _UpperCamelCase = { '''do_resize''': True, '''size''': {'''height''': 18, '''width''': 18}, '''do_normalize''': True, '''image_mean''': [0.5, 0.5, 0.5], '''image_std''': [0.5, 0.5, 0.5], } _UpperCamelCase = os.path.join(self.tmpdirname , _A ) with open(self.image_processor_file , '''w''' , encoding='''utf-8''' ) as fp: json.dump(_A , _A ) def UpperCamelCase_ ( self : Tuple , **_A : Optional[Any] ): return BertTokenizer.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : List[Any] , **_A : Union[str, Any] ): return ViTImageProcessor.from_pretrained(self.tmpdirname , **_A ) def UpperCamelCase_ ( self : int ): shutil.rmtree(self.tmpdirname ) def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = [np.random.randint(255 , size=(3, 30, 400) , dtype=np.uinta )] _UpperCamelCase = [Image.fromarray(np.moveaxis(_A , 0 , -1 ) ) for x in image_inputs] return image_inputs def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = self.get_image_processor() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) processor.save_pretrained(self.tmpdirname ) _UpperCamelCase = VisionTextDualEncoderProcessor.from_pretrained(self.tmpdirname ) self.assertEqual(processor.tokenizer.get_vocab() , tokenizer.get_vocab() ) self.assertIsInstance(processor.tokenizer , (BertTokenizer, BertTokenizerFast) ) self.assertEqual(processor.image_processor.to_json_string() , image_processor.to_json_string() ) self.assertIsInstance(processor.image_processor , _A ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = VisionTextDualEncoderProcessor( tokenizer=self.get_tokenizer() , image_processor=self.get_image_processor() ) processor.save_pretrained(self.tmpdirname ) _UpperCamelCase = self.get_tokenizer(bos_token='''(BOS)''' , eos_token='''(EOS)''' ) _UpperCamelCase = self.get_image_processor(do_normalize=_A , padding_value=1.0 ) _UpperCamelCase = VisionTextDualEncoderProcessor.from_pretrained( self.tmpdirname , bos_token='''(BOS)''' , eos_token='''(EOS)''' , do_normalize=_A , padding_value=1.0 ) self.assertEqual(processor.tokenizer.get_vocab() , tokenizer_add_kwargs.get_vocab() ) self.assertIsInstance(processor.tokenizer , (BertTokenizer, BertTokenizerFast) ) self.assertEqual(processor.image_processor.to_json_string() , image_processor_add_kwargs.to_json_string() ) self.assertIsInstance(processor.image_processor , _A ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = image_processor(_A , return_tensors='''np''' ) _UpperCamelCase = processor(images=_A , return_tensors='''np''' ) for key in input_feat_extract.keys(): self.assertAlmostEqual(input_feat_extract[key].sum() , input_processor[key].sum() , delta=1e-2 ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = processor(text=_A ) _UpperCamelCase = tokenizer(_A ) for key in encoded_tok.keys(): self.assertListEqual(encoded_tok[key] , encoded_processor[key] ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = processor(text=_A , images=_A ) self.assertListEqual(list(inputs.keys() ) , ['''input_ids''', '''token_type_ids''', '''attention_mask''', '''pixel_values'''] ) # test if it raises when no input is passed with self.assertRaises(_A ): processor() def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = [[1, 4, 5, 8, 1, 0, 8], [3, 4, 3, 1, 1, 8, 9]] _UpperCamelCase = processor.batch_decode(_A ) _UpperCamelCase = tokenizer.batch_decode(_A ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = self.get_image_processor() _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = VisionTextDualEncoderProcessor(tokenizer=_A , image_processor=_A ) _UpperCamelCase = '''lower newer''' _UpperCamelCase = self.prepare_image_inputs() _UpperCamelCase = processor(text=_A , images=_A ) self.assertListEqual(list(inputs.keys() ) , processor.model_input_names )
71
1
import unittest from typing import Tuple import torch from diffusers.utils import floats_tensor, randn_tensor, torch_all_close, torch_device from diffusers.utils.testing_utils import require_torch @require_torch class lowerCAmelCase_ : @property def UpperCamelCase_ ( self : Optional[int] ): return self.get_dummy_input() @property def UpperCamelCase_ ( self : Dict ): if self.block_type == "down": return (4, 32, 16, 16) elif self.block_type == "mid": return (4, 32, 32, 32) elif self.block_type == "up": return (4, 32, 64, 64) raise ValueError(F"""'{self.block_type}' is not a supported block_type. Set it to 'up', 'mid', or 'down'.""" ) def UpperCamelCase_ ( self : Union[str, Any] , _A : List[str]=True , _A : Any=False , _A : Union[str, Any]=False , _A : int=False , ): _UpperCamelCase = 4 _UpperCamelCase = 32 _UpperCamelCase = (32, 32) _UpperCamelCase = torch.manual_seed(0 ) _UpperCamelCase = torch.device(_A ) _UpperCamelCase = (batch_size, num_channels) + sizes _UpperCamelCase = randn_tensor(_A , generator=_A , device=_A ) _UpperCamelCase = {'''hidden_states''': hidden_states} if include_temb: _UpperCamelCase = 128 _UpperCamelCase = randn_tensor((batch_size, temb_channels) , generator=_A , device=_A ) if include_res_hidden_states_tuple: _UpperCamelCase = torch.manual_seed(1 ) _UpperCamelCase = (randn_tensor(_A , generator=_A , device=_A ),) if include_encoder_hidden_states: _UpperCamelCase = floats_tensor((batch_size, 32, 32) ).to(_A ) if include_skip_sample: _UpperCamelCase = randn_tensor(((batch_size, 3) + sizes) , generator=_A , device=_A ) return dummy_input def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = { '''in_channels''': 32, '''out_channels''': 32, '''temb_channels''': 128, } if self.block_type == "up": _UpperCamelCase = 32 if self.block_type == "mid": init_dict.pop('''out_channels''' ) _UpperCamelCase = self.dummy_input return init_dict, inputs_dict def UpperCamelCase_ ( self : Tuple , _A : Union[str, Any] ): _UpperCamelCase , _UpperCamelCase = self.prepare_init_args_and_inputs_for_common() _UpperCamelCase = self.block_class(**_A ) unet_block.to(_A ) unet_block.eval() with torch.no_grad(): _UpperCamelCase = unet_block(**_A ) if isinstance(_A , _A ): _UpperCamelCase = output[0] self.assertEqual(output.shape , self.output_shape ) _UpperCamelCase = output[0, -1, -3:, -3:] _UpperCamelCase = torch.tensor(_A ).to(_A ) assert torch_all_close(output_slice.flatten() , _A , atol=5e-3 ) @unittest.skipIf(torch_device == '''mps''' , '''Training is not supported in mps''' ) def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase , _UpperCamelCase = self.prepare_init_args_and_inputs_for_common() _UpperCamelCase = self.block_class(**_A ) model.to(_A ) model.train() _UpperCamelCase = model(**_A ) if isinstance(_A , _A ): _UpperCamelCase = output[0] _UpperCamelCase = torch.device(_A ) _UpperCamelCase = randn_tensor(output.shape , device=_A ) _UpperCamelCase = torch.nn.functional.mse_loss(_A , _A ) loss.backward()
71
def _snake_case ( __snake_case , __snake_case , __snake_case ): if n == 0: return 1 elif n % 2 == 1: return (binary_exponentiation(__snake_case , n - 1 , __snake_case ) * a) % mod else: _UpperCamelCase = binary_exponentiation(__snake_case , n / 2 , __snake_case ) return (b * b) % mod # a prime number _lowerCAmelCase = 701 _lowerCAmelCase = 1_000_000_000 _lowerCAmelCase = 10 # using binary exponentiation function, O(log(p)): print((a / b) % p == (a * binary_exponentiation(b, p - 2, p)) % p) print((a / b) % p == (a * b ** (p - 2)) % p)
71
1
import io import json import unittest from parameterized import parameterized from transformers import FSMTForConditionalGeneration, FSMTTokenizer from transformers.testing_utils import get_tests_dir, require_torch, slow, torch_device from utils import calculate_bleu _lowerCAmelCase = get_tests_dir() + "/test_data/fsmt/fsmt_val_data.json" with io.open(filename, "r", encoding="utf-8") as f: _lowerCAmelCase = json.load(f) @require_torch class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Optional[int] , _A : List[str] ): return FSMTTokenizer.from_pretrained(_A ) def UpperCamelCase_ ( self : str , _A : str ): _UpperCamelCase = FSMTForConditionalGeneration.from_pretrained(_A ).to(_A ) if torch_device == "cuda": model.half() return model @parameterized.expand( [ ['''en-ru''', 26.0], ['''ru-en''', 22.0], ['''en-de''', 22.0], ['''de-en''', 29.0], ] ) @slow def UpperCamelCase_ ( self : Union[str, Any] , _A : Any , _A : str ): # note: this test is not testing the best performance since it only evals a small batch # but it should be enough to detect a regression in the output quality _UpperCamelCase = F"""facebook/wmt19-{pair}""" _UpperCamelCase = self.get_tokenizer(_A ) _UpperCamelCase = self.get_model(_A ) _UpperCamelCase = bleu_data[pair]['''src'''] _UpperCamelCase = bleu_data[pair]['''tgt'''] _UpperCamelCase = tokenizer(_A , return_tensors='''pt''' , truncation=_A , padding='''longest''' ).to(_A ) _UpperCamelCase = model.generate( input_ids=batch.input_ids , num_beams=8 , ) _UpperCamelCase = tokenizer.batch_decode( _A , skip_special_tokens=_A , clean_up_tokenization_spaces=_A ) _UpperCamelCase = calculate_bleu(_A , _A ) print(_A ) self.assertGreaterEqual(scores['''bleu'''] , _A )
71
from math import cos, sin, sqrt, tau from audio_filters.iir_filter import IIRFilter def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 - _cos) / 2 _UpperCamelCase = 1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = (1 + _cos) / 2 _UpperCamelCase = -1 - _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = _sin / 2 _UpperCamelCase = 0 _UpperCamelCase = -ba _UpperCamelCase = 1 + alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 1 - alpha _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 + alpha _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([ba, ba, ba] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = 1 + alpha * big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha * big_a _UpperCamelCase = 1 + alpha / big_a _UpperCamelCase = -2 * _cos _UpperCamelCase = 1 - alpha / big_a _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (pmc + aaa) _UpperCamelCase = 2 * big_a * mpc _UpperCamelCase = big_a * (pmc - aaa) _UpperCamelCase = ppmc + aaa _UpperCamelCase = -2 * pmpc _UpperCamelCase = ppmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case = 1 / sqrt(2 ) , ): _UpperCamelCase = tau * frequency / samplerate _UpperCamelCase = sin(__snake_case ) _UpperCamelCase = cos(__snake_case ) _UpperCamelCase = _sin / (2 * q_factor) _UpperCamelCase = 10 ** (gain_db / 40) _UpperCamelCase = (big_a + 1) - (big_a - 1) * _cos _UpperCamelCase = (big_a + 1) + (big_a - 1) * _cos _UpperCamelCase = (big_a - 1) - (big_a + 1) * _cos _UpperCamelCase = (big_a - 1) + (big_a + 1) * _cos _UpperCamelCase = 2 * sqrt(__snake_case ) * alpha _UpperCamelCase = big_a * (ppmc + aaa) _UpperCamelCase = -2 * big_a * pmpc _UpperCamelCase = big_a * (ppmc - aaa) _UpperCamelCase = pmc + aaa _UpperCamelCase = 2 * mpc _UpperCamelCase = pmc - aaa _UpperCamelCase = IIRFilter(2 ) filt.set_coefficients([aa, aa, aa] , [ba, ba, ba] ) return filt
71
1
def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case , __snake_case , __snake_case ): if index == r: for j in range(__snake_case ): print(data[j] , end=''' ''' ) print(''' ''' ) return # When no more elements are there to put in data[] if i >= n: return # current is included, put next at next location _UpperCamelCase = arr[i] combination_util(__snake_case , __snake_case , __snake_case , index + 1 , __snake_case , i + 1 ) # current is excluded, replace it with # next (Note that i+1 is passed, but # index is not changed) combination_util(__snake_case , __snake_case , __snake_case , __snake_case , __snake_case , i + 1 ) # The main function that prints all combinations # of size r in arr[] of size n. This function # mainly uses combinationUtil() def _snake_case ( __snake_case , __snake_case , __snake_case ): # A temporary array to store all combination one by one _UpperCamelCase = [0] * r # Print all combination using temporary array 'data[]' combination_util(__snake_case , __snake_case , __snake_case , 0 , __snake_case , 0 ) if __name__ == "__main__": # Driver code to check the function above _lowerCAmelCase = [10, 20, 30, 40, 50] print_combination(arr, len(arr), 3) # This code is contributed by Ambuj sahu
71
from ...configuration_utils import PretrainedConfig from ...utils import logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = { "EleutherAI/gpt-neox-20b": "https://huggingface.co/EleutherAI/gpt-neox-20b/resolve/main/config.json", # See all GPTNeoX models at https://huggingface.co/models?filter=gpt_neox } class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "gpt_neox" def __init__( self : Union[str, Any] , _A : Union[str, Any]=5_0432 , _A : List[Any]=6144 , _A : int=44 , _A : int=64 , _A : Optional[Any]=2_4576 , _A : Any="gelu" , _A : Tuple=0.25 , _A : Union[str, Any]=1_0000 , _A : Tuple=0.0 , _A : Any=0.0 , _A : int=0.1 , _A : List[str]=2048 , _A : Dict=0.02 , _A : Optional[Any]=1e-5 , _A : Tuple=True , _A : List[Any]=0 , _A : Optional[int]=2 , _A : Optional[int]=False , _A : List[Any]=True , _A : Any=None , **_A : Any , ): super().__init__(bos_token_id=_A , eos_token_id=_A , **_A ) _UpperCamelCase = vocab_size _UpperCamelCase = max_position_embeddings _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = rotary_pct _UpperCamelCase = rotary_emb_base _UpperCamelCase = attention_dropout _UpperCamelCase = hidden_dropout _UpperCamelCase = classifier_dropout _UpperCamelCase = initializer_range _UpperCamelCase = layer_norm_eps _UpperCamelCase = use_cache _UpperCamelCase = tie_word_embeddings _UpperCamelCase = use_parallel_residual _UpperCamelCase = rope_scaling self._rope_scaling_validation() if self.hidden_size % self.num_attention_heads != 0: raise ValueError( '''The hidden size is not divisble by the number of attention heads! Make sure to update them!''' ) def UpperCamelCase_ ( self : str ): if self.rope_scaling is None: return if not isinstance(self.rope_scaling , _A ) or len(self.rope_scaling ) != 2: raise ValueError( '''`rope_scaling` must be a dictionary with with two fields, `name` and `factor`, ''' F"""got {self.rope_scaling}""" ) _UpperCamelCase = self.rope_scaling.get('''type''' , _A ) _UpperCamelCase = self.rope_scaling.get('''factor''' , _A ) if rope_scaling_type is None or rope_scaling_type not in ["linear", "dynamic"]: raise ValueError( F"""`rope_scaling`'s name field must be one of ['linear', 'dynamic'], got {rope_scaling_type}""" ) if rope_scaling_factor is None or not isinstance(_A , _A ) or rope_scaling_factor <= 1.0: raise ValueError(F"""`rope_scaling`'s factor field must be an float > 1, got {rope_scaling_factor}""" )
71
1
def _snake_case ( __snake_case , __snake_case , __snake_case ): return not any( neighbour == 1 and colored_vertices[i] == color for i, neighbour in enumerate(__snake_case ) ) def _snake_case ( __snake_case , __snake_case , __snake_case , __snake_case ): # Base Case if index == len(__snake_case ): return True # Recursive Step for i in range(__snake_case ): if valid_coloring(graph[index] , __snake_case , __snake_case ): # Color current vertex _UpperCamelCase = i # Validate coloring if util_color(__snake_case , __snake_case , __snake_case , index + 1 ): return True # Backtrack _UpperCamelCase = -1 return False def _snake_case ( __snake_case , __snake_case ): _UpperCamelCase = [-1] * len(__snake_case ) if util_color(__snake_case , __snake_case , __snake_case , 0 ): return colored_vertices return []
71
from ..utils import DummyObject, requires_backends class lowerCAmelCase_ ( metaclass=__lowercase ): UpperCAmelCase = ["keras_nlp"] def __init__( self : Any , *_A : Dict , **_A : List[str] ): requires_backends(self , ['''keras_nlp'''] )
71
1
import unittest from transformers import load_tool from .test_tools_common import ToolTesterMixin _lowerCAmelCase = "\nHugging Face was founded in 2016 by French entrepreneurs Clément Delangue, Julien Chaumond, and Thomas Wolf originally as a company that developed a chatbot app targeted at teenagers.[2] After open-sourcing the model behind the chatbot, the company pivoted to focus on being a platform for machine learning.\n\nIn March 2021, Hugging Face raised $40 million in a Series B funding round.[3]\n\nOn April 28, 2021, the company launched the BigScience Research Workshop in collaboration with several other research groups to release an open large language model.[4] In 2022, the workshop concluded with the announcement of BLOOM, a multilingual large language model with 176 billion parameters.[5]\n" class lowerCAmelCase_ ( unittest.TestCase, __lowercase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = load_tool('''text-question-answering''' ) self.tool.setup() _UpperCamelCase = load_tool('''text-question-answering''' , remote=_A ) def UpperCamelCase_ ( self : Any ): _UpperCamelCase = self.tool(_A , '''What did Hugging Face do in April 2021?''' ) self.assertEqual(_A , '''launched the BigScience Research Workshop''' ) def UpperCamelCase_ ( self : List[Any] ): _UpperCamelCase = self.remote_tool(_A , '''What did Hugging Face do in April 2021?''' ) self.assertEqual(_A , '''launched the BigScience Research Workshop''' ) def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = self.tool(text=_A , question='''What did Hugging Face do in April 2021?''' ) self.assertEqual(_A , '''launched the BigScience Research Workshop''' ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.remote_tool(text=_A , question='''What did Hugging Face do in April 2021?''' ) self.assertEqual(_A , '''launched the BigScience Research Workshop''' )
71
from typing import Optional, Tuple, Union import tensorflow as tf from ...activations_tf import ACTaFN from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward from ...modeling_tf_outputs import ( TFBaseModelOutputWithNoAttention, TFBaseModelOutputWithPoolingAndNoAttention, TFSequenceClassifierOutput, ) from ...modeling_tf_utils import TFPreTrainedModel, TFSequenceClassificationLoss, keras_serializable, unpack_inputs from ...tf_utils import shape_list from ...utils import logging from .configuration_regnet import RegNetConfig _lowerCAmelCase = logging.get_logger(__name__) # General docstring _lowerCAmelCase = "RegNetConfig" # Base docstring _lowerCAmelCase = "facebook/regnet-y-040" _lowerCAmelCase = [1, 1_088, 7, 7] # Image classification docstring _lowerCAmelCase = "facebook/regnet-y-040" _lowerCAmelCase = "tabby, tabby cat" _lowerCAmelCase = [ "facebook/regnet-y-040", # See all regnet models at https://huggingface.co/models?filter=regnet ] class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : str , _A : int , _A : int = 3 , _A : int = 1 , _A : int = 1 , _A : Optional[str] = "relu" , **_A : Any , ): super().__init__(**_A ) # The padding and conv has been verified in # https://colab.research.google.com/gist/sayakpaul/854bc10eeaf21c9ee2119e0b9f3841a7/scratchpad.ipynb _UpperCamelCase = tf.keras.layers.ZeroPaddingaD(padding=kernel_size // 2 ) _UpperCamelCase = tf.keras.layers.ConvaD( filters=_A , kernel_size=_A , strides=_A , padding='''VALID''' , groups=_A , use_bias=_A , name='''convolution''' , ) _UpperCamelCase = tf.keras.layers.BatchNormalization(epsilon=1e-5 , momentum=0.9 , name='''normalization''' ) _UpperCamelCase = ACTaFN[activation] if activation is not None else tf.identity def UpperCamelCase_ ( self : Any , _A : Any ): _UpperCamelCase = self.convolution(self.padding(_A ) ) _UpperCamelCase = self.normalization(_A ) _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Optional[Any] , _A : RegNetConfig , **_A : Any ): super().__init__(**_A ) _UpperCamelCase = config.num_channels _UpperCamelCase = TFRegNetConvLayer( out_channels=config.embedding_size , kernel_size=3 , stride=2 , activation=config.hidden_act , name='''embedder''' , ) def UpperCamelCase_ ( self : List[str] , _A : Optional[int] ): _UpperCamelCase = shape_list(_A )[1] if tf.executing_eagerly() and num_channels != self.num_channels: raise ValueError( '''Make sure that the channel dimension of the pixel values match with the one set in the configuration.''' ) # When running on CPU, `tf.keras.layers.Conv2D` doesn't support `NCHW` format. # So change the input format from `NCHW` to `NHWC`. # shape = (batch_size, in_height, in_width, in_channels=num_channels) _UpperCamelCase = tf.transpose(_A , perm=(0, 2, 3, 1) ) _UpperCamelCase = self.embedder(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : str , _A : int , _A : int = 2 , **_A : Optional[Any] ): super().__init__(**_A ) _UpperCamelCase = tf.keras.layers.ConvaD( filters=_A , kernel_size=1 , strides=_A , use_bias=_A , name='''convolution''' ) _UpperCamelCase = tf.keras.layers.BatchNormalization(epsilon=1e-5 , momentum=0.9 , name='''normalization''' ) def UpperCamelCase_ ( self : str , _A : tf.Tensor , _A : bool = False ): return self.normalization(self.convolution(_A ) , training=_A ) class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Dict , _A : int , _A : int , **_A : Dict ): super().__init__(**_A ) _UpperCamelCase = tf.keras.layers.GlobalAveragePoolingaD(keepdims=_A , name='''pooler''' ) _UpperCamelCase = [ tf.keras.layers.ConvaD(filters=_A , kernel_size=1 , activation='''relu''' , name='''attention.0''' ), tf.keras.layers.ConvaD(filters=_A , kernel_size=1 , activation='''sigmoid''' , name='''attention.2''' ), ] def UpperCamelCase_ ( self : List[str] , _A : List[Any] ): # [batch_size, h, w, num_channels] -> [batch_size, 1, 1, num_channels] _UpperCamelCase = self.pooler(_A ) for layer_module in self.attention: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = hidden_state * pooled return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , _A : int , _A : int , _A : int = 1 , **_A : str ): super().__init__(**_A ) _UpperCamelCase = in_channels != out_channels or stride != 1 _UpperCamelCase = max(1 , out_channels // config.groups_width ) _UpperCamelCase = ( TFRegNetShortCut(_A , stride=_A , name='''shortcut''' ) if should_apply_shortcut else tf.keras.layers.Activation('''linear''' , name='''shortcut''' ) ) # `self.layers` instead of `self.layer` because that is a reserved argument. _UpperCamelCase = [ TFRegNetConvLayer(_A , kernel_size=1 , activation=config.hidden_act , name='''layer.0''' ), TFRegNetConvLayer( _A , stride=_A , groups=_A , activation=config.hidden_act , name='''layer.1''' ), TFRegNetConvLayer(_A , kernel_size=1 , activation=_A , name='''layer.2''' ), ] _UpperCamelCase = ACTaFN[config.hidden_act] def UpperCamelCase_ ( self : Dict , _A : Tuple ): _UpperCamelCase = hidden_state for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = self.shortcut(_A ) hidden_state += residual _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , _A : int , _A : int , _A : int = 1 , **_A : int ): super().__init__(**_A ) _UpperCamelCase = in_channels != out_channels or stride != 1 _UpperCamelCase = max(1 , out_channels // config.groups_width ) _UpperCamelCase = ( TFRegNetShortCut(_A , stride=_A , name='''shortcut''' ) if should_apply_shortcut else tf.keras.layers.Activation('''linear''' , name='''shortcut''' ) ) _UpperCamelCase = [ TFRegNetConvLayer(_A , kernel_size=1 , activation=config.hidden_act , name='''layer.0''' ), TFRegNetConvLayer( _A , stride=_A , groups=_A , activation=config.hidden_act , name='''layer.1''' ), TFRegNetSELayer(_A , reduced_channels=int(round(in_channels / 4 ) ) , name='''layer.2''' ), TFRegNetConvLayer(_A , kernel_size=1 , activation=_A , name='''layer.3''' ), ] _UpperCamelCase = ACTaFN[config.hidden_act] def UpperCamelCase_ ( self : Tuple , _A : List[Any] ): _UpperCamelCase = hidden_state for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) _UpperCamelCase = self.shortcut(_A ) hidden_state += residual _UpperCamelCase = self.activation(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : Tuple , _A : RegNetConfig , _A : int , _A : int , _A : int = 2 , _A : int = 2 , **_A : Union[str, Any] ): super().__init__(**_A ) _UpperCamelCase = TFRegNetXLayer if config.layer_type == '''x''' else TFRegNetYLayer _UpperCamelCase = [ # downsampling is done in the first layer with stride of 2 layer(_A , _A , _A , stride=_A , name='''layers.0''' ), *[layer(_A , _A , _A , name=F"""layers.{i+1}""" ) for i in range(depth - 1 )], ] def UpperCamelCase_ ( self : Union[str, Any] , _A : Optional[int] ): for layer_module in self.layers: _UpperCamelCase = layer_module(_A ) return hidden_state class lowerCAmelCase_ ( tf.keras.layers.Layer ): def __init__( self : List[Any] , _A : RegNetConfig , **_A : List[str] ): super().__init__(**_A ) _UpperCamelCase = [] # based on `downsample_in_first_stage`, the first layer of the first stage may or may not downsample the input self.stages.append( TFRegNetStage( _A , config.embedding_size , config.hidden_sizes[0] , stride=2 if config.downsample_in_first_stage else 1 , depth=config.depths[0] , name='''stages.0''' , ) ) _UpperCamelCase = zip(config.hidden_sizes , config.hidden_sizes[1:] ) for i, ((in_channels, out_channels), depth) in enumerate(zip(_A , config.depths[1:] ) ): self.stages.append(TFRegNetStage(_A , _A , _A , depth=_A , name=F"""stages.{i+1}""" ) ) def UpperCamelCase_ ( self : Optional[int] , _A : tf.Tensor , _A : bool = False , _A : bool = True ): _UpperCamelCase = () if output_hidden_states else None for stage_module in self.stages: if output_hidden_states: _UpperCamelCase = hidden_states + (hidden_state,) _UpperCamelCase = stage_module(_A ) if output_hidden_states: _UpperCamelCase = hidden_states + (hidden_state,) if not return_dict: return tuple(v for v in [hidden_state, hidden_states] if v is not None ) return TFBaseModelOutputWithNoAttention(last_hidden_state=_A , hidden_states=_A ) @keras_serializable class lowerCAmelCase_ ( tf.keras.layers.Layer ): UpperCAmelCase = RegNetConfig def __init__( self : int , _A : Tuple , **_A : int ): super().__init__(**_A ) _UpperCamelCase = config _UpperCamelCase = TFRegNetEmbeddings(_A , name='''embedder''' ) _UpperCamelCase = TFRegNetEncoder(_A , name='''encoder''' ) _UpperCamelCase = tf.keras.layers.GlobalAveragePoolingaD(keepdims=_A , name='''pooler''' ) @unpack_inputs def UpperCamelCase_ ( self : Optional[int] , _A : tf.Tensor , _A : Optional[bool] = None , _A : Optional[bool] = None , _A : bool = False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.embedder(_A , training=_A ) _UpperCamelCase = self.encoder( _A , output_hidden_states=_A , return_dict=_A , training=_A ) _UpperCamelCase = encoder_outputs[0] _UpperCamelCase = self.pooler(_A ) # Change to NCHW output format have uniformity in the modules _UpperCamelCase = tf.transpose(_A , perm=(0, 3, 1, 2) ) _UpperCamelCase = tf.transpose(_A , perm=(0, 3, 1, 2) ) # Change the other hidden state outputs to NCHW as well if output_hidden_states: _UpperCamelCase = tuple([tf.transpose(_A , perm=(0, 3, 1, 2) ) for h in encoder_outputs[1]] ) if not return_dict: return (last_hidden_state, pooled_output) + encoder_outputs[1:] return TFBaseModelOutputWithPoolingAndNoAttention( last_hidden_state=_A , pooler_output=_A , hidden_states=hidden_states if output_hidden_states else encoder_outputs.hidden_states , ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = RegNetConfig UpperCAmelCase = "regnet" UpperCAmelCase = "pixel_values" @property def UpperCamelCase_ ( self : Tuple ): return {"pixel_values": tf.TensorSpec(shape=(None, self.config.num_channels, 224, 224) , dtype=tf.floataa )} _lowerCAmelCase = r"\n Parameters:\n This model is a Tensorflow\n [tf.keras.layers.Layer](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Layer) sub-class. Use it as a\n regular Tensorflow Module and refer to the Tensorflow documentation for all matter related to general usage and\n behavior.\n config ([`RegNetConfig`]): Model configuration class with all the parameters of the model.\n Initializing with a config file does not load the weights associated with the model, only the\n configuration. Check out the [`~TFPreTrainedModel.from_pretrained`] method to load the model weights.\n" _lowerCAmelCase = r"\n Args:\n pixel_values (`tf.Tensor` of shape `(batch_size, num_channels, height, width)`):\n Pixel values. Pixel values can be obtained using [`AutoImageProcessor`]. See\n [`ConveNextImageProcessor.__call__`] for details.\n output_hidden_states (`bool`, *optional*):\n Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for\n more detail.\n return_dict (`bool`, *optional*):\n Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.\n" @add_start_docstrings( "The bare RegNet model outputting raw features without any specific head on top.", __lowercase, ) class lowerCAmelCase_ ( __lowercase ): def __init__( self : List[Any] , _A : RegNetConfig , *_A : Optional[int] , **_A : Tuple ): super().__init__(_A , *_A , **_A ) _UpperCamelCase = TFRegNetMainLayer(_A , name='''regnet''' ) @unpack_inputs @add_start_docstrings_to_model_forward(_A ) @add_code_sample_docstrings( checkpoint=_CHECKPOINT_FOR_DOC , output_type=_A , config_class=_CONFIG_FOR_DOC , modality='''vision''' , expected_output=_EXPECTED_OUTPUT_SHAPE , ) def UpperCamelCase_ ( self : Any , _A : tf.Tensor , _A : Optional[bool] = None , _A : Optional[bool] = None , _A : Optional[int]=False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.regnet( pixel_values=_A , output_hidden_states=_A , return_dict=_A , training=_A , ) if not return_dict: return (outputs[0],) + outputs[1:] return TFBaseModelOutputWithPoolingAndNoAttention( last_hidden_state=outputs.last_hidden_state , pooler_output=outputs.pooler_output , hidden_states=outputs.hidden_states , ) @add_start_docstrings( "\n RegNet Model with an image classification head on top (a linear layer on top of the pooled features), e.g. for\n ImageNet.\n ", __lowercase, ) class lowerCAmelCase_ ( __lowercase, __lowercase ): def __init__( self : List[Any] , _A : RegNetConfig , *_A : Any , **_A : int ): super().__init__(_A , *_A , **_A ) _UpperCamelCase = config.num_labels _UpperCamelCase = TFRegNetMainLayer(_A , name='''regnet''' ) # classification head _UpperCamelCase = [ tf.keras.layers.Flatten(), tf.keras.layers.Dense(config.num_labels , name='''classifier.1''' ) if config.num_labels > 0 else tf.identity, ] @unpack_inputs @add_start_docstrings_to_model_forward(_A ) @add_code_sample_docstrings( checkpoint=_IMAGE_CLASS_CHECKPOINT , output_type=_A , config_class=_CONFIG_FOR_DOC , expected_output=_IMAGE_CLASS_EXPECTED_OUTPUT , ) def UpperCamelCase_ ( self : str , _A : tf.Tensor = None , _A : tf.Tensor = None , _A : bool = None , _A : bool = None , _A : Any=False , ): _UpperCamelCase = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) _UpperCamelCase = return_dict if return_dict is not None else self.config.use_return_dict _UpperCamelCase = self.regnet( _A , output_hidden_states=_A , return_dict=_A , training=_A ) _UpperCamelCase = outputs.pooler_output if return_dict else outputs[1] _UpperCamelCase = self.classifier[0](_A ) _UpperCamelCase = self.classifier[1](_A ) _UpperCamelCase = None if labels is None else self.hf_compute_loss(labels=_A , logits=_A ) if not return_dict: _UpperCamelCase = (logits,) + outputs[2:] return ((loss,) + output) if loss is not None else output return TFSequenceClassifierOutput(loss=_A , logits=_A , hidden_states=outputs.hidden_states )
71
1
import math from dataclasses import dataclass from typing import Optional, Tuple, Union import numpy as np import torch from ..configuration_utils import ConfigMixin, register_to_config from ..utils import BaseOutput, randn_tensor from .scheduling_utils import SchedulerMixin @dataclass # Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->UnCLIP class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = 42 UpperCAmelCase = None def _snake_case ( __snake_case , __snake_case=0.999 , __snake_case="cosine" , ): if alpha_transform_type == "cosine": def alpha_bar_fn(__snake_case ): return math.cos((t + 0.008) / 1.008 * math.pi / 2 ) ** 2 elif alpha_transform_type == "exp": def alpha_bar_fn(__snake_case ): return math.exp(t * -12.0 ) else: raise ValueError(f"""Unsupported alpha_tranform_type: {alpha_transform_type}""" ) _UpperCamelCase = [] for i in range(__snake_case ): _UpperCamelCase = i / num_diffusion_timesteps _UpperCamelCase = (i + 1) / num_diffusion_timesteps betas.append(min(1 - alpha_bar_fn(__snake_case ) / alpha_bar_fn(__snake_case ) , __snake_case ) ) return torch.tensor(__snake_case , dtype=torch.floataa ) class lowerCAmelCase_ ( __lowercase, __lowercase ): @register_to_config def __init__( self : Optional[int] , _A : int = 1000 , _A : str = "fixed_small_log" , _A : bool = True , _A : Optional[float] = 1.0 , _A : str = "epsilon" , _A : str = "squaredcos_cap_v2" , ): if beta_schedule != "squaredcos_cap_v2": raise ValueError('''UnCLIPScheduler only supports `beta_schedule`: \'squaredcos_cap_v2\'''' ) _UpperCamelCase = betas_for_alpha_bar(_A ) _UpperCamelCase = 1.0 - self.betas _UpperCamelCase = torch.cumprod(self.alphas , dim=0 ) _UpperCamelCase = torch.tensor(1.0 ) # standard deviation of the initial noise distribution _UpperCamelCase = 1.0 # setable values _UpperCamelCase = None _UpperCamelCase = torch.from_numpy(np.arange(0 , _A )[::-1].copy() ) _UpperCamelCase = variance_type def UpperCamelCase_ ( self : List[Any] , _A : torch.FloatTensor , _A : Optional[int] = None ): return sample def UpperCamelCase_ ( self : List[Any] , _A : int , _A : Union[str, torch.device] = None ): _UpperCamelCase = num_inference_steps _UpperCamelCase = (self.config.num_train_timesteps - 1) / (self.num_inference_steps - 1) _UpperCamelCase = (np.arange(0 , _A ) * step_ratio).round()[::-1].copy().astype(np.intaa ) _UpperCamelCase = torch.from_numpy(_A ).to(_A ) def UpperCamelCase_ ( self : Dict , _A : Tuple , _A : str=None , _A : Tuple=None , _A : List[str]=None ): if prev_timestep is None: _UpperCamelCase = t - 1 _UpperCamelCase = self.alphas_cumprod[t] _UpperCamelCase = self.alphas_cumprod[prev_timestep] if prev_timestep >= 0 else self.one _UpperCamelCase = 1 - alpha_prod_t _UpperCamelCase = 1 - alpha_prod_t_prev if prev_timestep == t - 1: _UpperCamelCase = self.betas[t] else: _UpperCamelCase = 1 - alpha_prod_t / alpha_prod_t_prev # For t > 0, compute predicted variance βt (see formula (6) and (7) from https://arxiv.org/pdf/2006.11239.pdf) # and sample from it to get previous sample # x_{t-1} ~ N(pred_prev_sample, variance) == add variance to pred_sample _UpperCamelCase = beta_prod_t_prev / beta_prod_t * beta if variance_type is None: _UpperCamelCase = self.config.variance_type # hacks - were probably added for training stability if variance_type == "fixed_small_log": _UpperCamelCase = torch.log(torch.clamp(_A , min=1e-20 ) ) _UpperCamelCase = torch.exp(0.5 * variance ) elif variance_type == "learned_range": # NOTE difference with DDPM scheduler _UpperCamelCase = variance.log() _UpperCamelCase = beta.log() _UpperCamelCase = (predicted_variance + 1) / 2 _UpperCamelCase = frac * max_log + (1 - frac) * min_log return variance def UpperCamelCase_ ( self : Optional[int] , _A : torch.FloatTensor , _A : int , _A : torch.FloatTensor , _A : Optional[int] = None , _A : str=None , _A : bool = True , ): _UpperCamelCase = timestep if model_output.shape[1] == sample.shape[1] * 2 and self.variance_type == "learned_range": _UpperCamelCase , _UpperCamelCase = torch.split(_A , sample.shape[1] , dim=1 ) else: _UpperCamelCase = None # 1. compute alphas, betas if prev_timestep is None: _UpperCamelCase = t - 1 _UpperCamelCase = self.alphas_cumprod[t] _UpperCamelCase = self.alphas_cumprod[prev_timestep] if prev_timestep >= 0 else self.one _UpperCamelCase = 1 - alpha_prod_t _UpperCamelCase = 1 - alpha_prod_t_prev if prev_timestep == t - 1: _UpperCamelCase = self.betas[t] _UpperCamelCase = self.alphas[t] else: _UpperCamelCase = 1 - alpha_prod_t / alpha_prod_t_prev _UpperCamelCase = 1 - beta # 2. compute predicted original sample from predicted noise also called # "predicted x_0" of formula (15) from https://arxiv.org/pdf/2006.11239.pdf if self.config.prediction_type == "epsilon": _UpperCamelCase = (sample - beta_prod_t ** 0.5 * model_output) / alpha_prod_t ** 0.5 elif self.config.prediction_type == "sample": _UpperCamelCase = model_output else: raise ValueError( F"""prediction_type given as {self.config.prediction_type} must be one of `epsilon` or `sample`""" ''' for the UnCLIPScheduler.''' ) # 3. Clip "predicted x_0" if self.config.clip_sample: _UpperCamelCase = torch.clamp( _A , -self.config.clip_sample_range , self.config.clip_sample_range ) # 4. Compute coefficients for pred_original_sample x_0 and current sample x_t # See formula (7) from https://arxiv.org/pdf/2006.11239.pdf _UpperCamelCase = (alpha_prod_t_prev ** 0.5 * beta) / beta_prod_t _UpperCamelCase = alpha ** 0.5 * beta_prod_t_prev / beta_prod_t # 5. Compute predicted previous sample µ_t # See formula (7) from https://arxiv.org/pdf/2006.11239.pdf _UpperCamelCase = pred_original_sample_coeff * pred_original_sample + current_sample_coeff * sample # 6. Add noise _UpperCamelCase = 0 if t > 0: _UpperCamelCase = randn_tensor( model_output.shape , dtype=model_output.dtype , generator=_A , device=model_output.device ) _UpperCamelCase = self._get_variance( _A , predicted_variance=_A , prev_timestep=_A , ) if self.variance_type == "fixed_small_log": _UpperCamelCase = variance elif self.variance_type == "learned_range": _UpperCamelCase = (0.5 * variance).exp() else: raise ValueError( F"""variance_type given as {self.variance_type} must be one of `fixed_small_log` or `learned_range`""" ''' for the UnCLIPScheduler.''' ) _UpperCamelCase = variance * variance_noise _UpperCamelCase = pred_prev_sample + variance if not return_dict: return (pred_prev_sample,) return UnCLIPSchedulerOutput(prev_sample=_A , pred_original_sample=_A ) def UpperCamelCase_ ( self : Optional[Any] , _A : torch.FloatTensor , _A : torch.FloatTensor , _A : torch.IntTensor , ): # Make sure alphas_cumprod and timestep have same device and dtype as original_samples _UpperCamelCase = self.alphas_cumprod.to(device=original_samples.device , dtype=original_samples.dtype ) _UpperCamelCase = timesteps.to(original_samples.device ) _UpperCamelCase = alphas_cumprod[timesteps] ** 0.5 _UpperCamelCase = sqrt_alpha_prod.flatten() while len(sqrt_alpha_prod.shape ) < len(original_samples.shape ): _UpperCamelCase = sqrt_alpha_prod.unsqueeze(-1 ) _UpperCamelCase = (1 - alphas_cumprod[timesteps]) ** 0.5 _UpperCamelCase = sqrt_one_minus_alpha_prod.flatten() while len(sqrt_one_minus_alpha_prod.shape ) < len(original_samples.shape ): _UpperCamelCase = sqrt_one_minus_alpha_prod.unsqueeze(-1 ) _UpperCamelCase = sqrt_alpha_prod * original_samples + sqrt_one_minus_alpha_prod * noise return noisy_samples
71
from sklearn.metrics import mean_squared_error import datasets _lowerCAmelCase = "\\n@article{scikit-learn,\n title={Scikit-learn: Machine Learning in {P}ython},\n author={Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V.\n and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P.\n and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and\n Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E.},\n journal={Journal of Machine Learning Research},\n volume={12},\n pages={2825--2830},\n year={2011}\n}\n" _lowerCAmelCase = "\\nMean Squared Error(MSE) is the average of the square of difference between the predicted\nand actual values.\n" _lowerCAmelCase = "\nArgs:\n predictions: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Estimated target values.\n references: array-like of shape (n_samples,) or (n_samples, n_outputs)\n Ground truth (correct) target values.\n sample_weight: array-like of shape (n_samples,), default=None\n Sample weights.\n multioutput: {\"raw_values\", \"uniform_average\"} or array-like of shape (n_outputs,), default=\"uniform_average\"\n Defines aggregating of multiple output values. Array-like value defines weights used to average errors.\n\n \"raw_values\" : Returns a full set of errors in case of multioutput input.\n\n \"uniform_average\" : Errors of all outputs are averaged with uniform weight.\n\n squared : bool, default=True\n If True returns MSE value, if False returns RMSE (Root Mean Squared Error) value.\n\nReturns:\n mse : mean squared error.\nExamples:\n\n >>> mse_metric = datasets.load_metric(\"mse\")\n >>> predictions = [2.5, 0.0, 2, 8]\n >>> references = [3, -0.5, 2, 7]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.375}\n >>> rmse_result = mse_metric.compute(predictions=predictions, references=references, squared=False)\n >>> print(rmse_result)\n {'mse': 0.6123724356957945}\n\n If you're using multi-dimensional lists, then set the config as follows :\n\n >>> mse_metric = datasets.load_metric(\"mse\", \"multilist\")\n >>> predictions = [[0.5, 1], [-1, 1], [7, -6]]\n >>> references = [[0, 2], [-1, 2], [8, -5]]\n >>> results = mse_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {'mse': 0.7083333333333334}\n >>> results = mse_metric.compute(predictions=predictions, references=references, multioutput='raw_values')\n >>> print(results) # doctest: +NORMALIZE_WHITESPACE\n {'mse': array([0.41666667, 1. ])}\n" @datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION ) class lowerCAmelCase_ ( datasets.Metric ): def UpperCamelCase_ ( self : Optional[int] ): return datasets.MetricInfo( description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features(self._get_feature_types() ) , reference_urls=[ '''https://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html''' ] , ) def UpperCamelCase_ ( self : Dict ): if self.config_name == "multilist": return { "predictions": datasets.Sequence(datasets.Value('''float''' ) ), "references": datasets.Sequence(datasets.Value('''float''' ) ), } else: return { "predictions": datasets.Value('''float''' ), "references": datasets.Value('''float''' ), } def UpperCamelCase_ ( self : Any , _A : List[Any] , _A : List[str] , _A : Dict=None , _A : List[str]="uniform_average" , _A : int=True ): _UpperCamelCase = mean_squared_error( _A , _A , sample_weight=_A , multioutput=_A , squared=_A ) return {"mse": mse}
71
1
import unittest from transformers import ( MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING, TF_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING, TextaTextGenerationPipeline, pipeline, ) from transformers.testing_utils import is_pipeline_test, require_tf, require_torch from transformers.utils import is_torch_available from .test_pipelines_common import ANY if is_torch_available(): import torch @is_pipeline_test class lowerCAmelCase_ ( unittest.TestCase ): UpperCAmelCase = MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING UpperCAmelCase = TF_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING def UpperCamelCase_ ( self : Dict , _A : Union[str, Any] , _A : str , _A : Any ): _UpperCamelCase = TextaTextGenerationPipeline(model=_A , tokenizer=_A ) return generator, ["Something to write", "Something else"] def UpperCamelCase_ ( self : str , _A : Union[str, Any] , _A : Optional[int] ): _UpperCamelCase = generator('''Something there''' ) self.assertEqual(_A , [{'''generated_text''': ANY(_A )}] ) # These are encoder decoder, they don't just append to incoming string self.assertFalse(outputs[0]['''generated_text'''].startswith('''Something there''' ) ) _UpperCamelCase = generator(['''This is great !''', '''Something else'''] , num_return_sequences=2 , do_sample=_A ) self.assertEqual( _A , [ [{'''generated_text''': ANY(_A )}, {'''generated_text''': ANY(_A )}], [{'''generated_text''': ANY(_A )}, {'''generated_text''': ANY(_A )}], ] , ) _UpperCamelCase = generator( ['''This is great !''', '''Something else'''] , num_return_sequences=2 , batch_size=2 , do_sample=_A ) self.assertEqual( _A , [ [{'''generated_text''': ANY(_A )}, {'''generated_text''': ANY(_A )}], [{'''generated_text''': ANY(_A )}, {'''generated_text''': ANY(_A )}], ] , ) with self.assertRaises(_A ): generator(4 ) @require_torch def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = pipeline('''text2text-generation''' , model='''patrickvonplaten/t5-tiny-random''' , framework='''pt''' ) # do_sample=False necessary for reproducibility _UpperCamelCase = generator('''Something there''' , do_sample=_A ) self.assertEqual(_A , [{'''generated_text''': ''''''}] ) _UpperCamelCase = 3 _UpperCamelCase = generator( '''Something there''' , num_return_sequences=_A , num_beams=_A , ) _UpperCamelCase = [ {'''generated_text''': '''Beide Beide Beide Beide Beide Beide Beide Beide Beide'''}, {'''generated_text''': '''Beide Beide Beide Beide Beide Beide Beide Beide'''}, {'''generated_text''': ''''''}, ] self.assertEqual(_A , _A ) _UpperCamelCase = generator('''This is a test''' , do_sample=_A , num_return_sequences=2 , return_tensors=_A ) self.assertEqual( _A , [ {'''generated_token_ids''': ANY(torch.Tensor )}, {'''generated_token_ids''': ANY(torch.Tensor )}, ] , ) _UpperCamelCase = generator.model.config.eos_token_id _UpperCamelCase = '''<pad>''' _UpperCamelCase = generator( ['''This is a test''', '''This is a second test'''] , do_sample=_A , num_return_sequences=2 , batch_size=2 , return_tensors=_A , ) self.assertEqual( _A , [ [ {'''generated_token_ids''': ANY(torch.Tensor )}, {'''generated_token_ids''': ANY(torch.Tensor )}, ], [ {'''generated_token_ids''': ANY(torch.Tensor )}, {'''generated_token_ids''': ANY(torch.Tensor )}, ], ] , ) @require_tf def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = pipeline('''text2text-generation''' , model='''patrickvonplaten/t5-tiny-random''' , framework='''tf''' ) # do_sample=False necessary for reproducibility _UpperCamelCase = generator('''Something there''' , do_sample=_A ) self.assertEqual(_A , [{'''generated_text''': ''''''}] )
71
import os import re import shutil import sys import tempfile import unittest import black _lowerCAmelCase = os.path.abspath(os.path.dirname(os.path.dirname(os.path.dirname(__file__)))) sys.path.append(os.path.join(git_repo_path, "utils")) import check_copies # noqa: E402 # This is the reference code that will be used in the tests. # If DDPMSchedulerOutput is changed in scheduling_ddpm.py, this code needs to be manually updated. _lowerCAmelCase = " \"\"\"\n Output class for the scheduler's step function output.\n\n Args:\n prev_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n Computed sample (x_{t-1}) of previous timestep. `prev_sample` should be used as next model input in the\n denoising loop.\n pred_original_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):\n The predicted denoised sample (x_{0}) based on the model output from the current timestep.\n `pred_original_sample` can be used to preview progress or for guidance.\n \"\"\"\n\n prev_sample: torch.FloatTensor\n pred_original_sample: Optional[torch.FloatTensor] = None\n" class lowerCAmelCase_ ( unittest.TestCase ): def UpperCamelCase_ ( self : Any ): _UpperCamelCase = tempfile.mkdtemp() os.makedirs(os.path.join(self.diffusers_dir , '''schedulers/''' ) ) _UpperCamelCase = self.diffusers_dir shutil.copy( os.path.join(_A , '''src/diffusers/schedulers/scheduling_ddpm.py''' ) , os.path.join(self.diffusers_dir , '''schedulers/scheduling_ddpm.py''' ) , ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = '''src/diffusers''' shutil.rmtree(self.diffusers_dir ) def UpperCamelCase_ ( self : Union[str, Any] , _A : Tuple , _A : Optional[Any] , _A : Dict , _A : List[str]=None ): _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + class_code if overwrite_result is not None: _UpperCamelCase = comment + F"""\nclass {class_name}(nn.Module):\n""" + overwrite_result _UpperCamelCase = black.Mode(target_versions={black.TargetVersion.PYaa} , line_length=119 ) _UpperCamelCase = black.format_str(_A , mode=_A ) _UpperCamelCase = os.path.join(self.diffusers_dir , '''new_code.py''' ) with open(_A , '''w''' , newline='''\n''' ) as f: f.write(_A ) if overwrite_result is None: self.assertTrue(len(check_copies.is_copy_consistent(_A ) ) == 0 ) else: check_copies.is_copy_consistent(f.name , overwrite=_A ) with open(_A , '''r''' ) as f: self.assertTrue(f.read() , _A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = check_copies.find_code_in_diffusers('''schedulers.scheduling_ddpm.DDPMSchedulerOutput''' ) self.assertEqual(_A , _A ) def UpperCamelCase_ ( self : Optional[Any] ): # Base copy consistency self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , REFERENCE_CODE + '''\n''' , ) # With no empty line at the end self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput''' , '''DDPMSchedulerOutput''' , _A , ) # Copy consistency with rename self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , re.sub('''DDPM''' , '''Test''' , _A ) , ) # Copy consistency with a really long name _UpperCamelCase = '''TestClassWithAReallyLongNameBecauseSomePeopleLikeThatForSomeReason''' self.check_copy_consistency( F"""# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->{long_class_name}""" , F"""{long_class_name}SchedulerOutput""" , re.sub('''Bert''' , _A , _A ) , ) # Copy consistency with overwrite self.check_copy_consistency( '''# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->Test''' , '''TestSchedulerOutput''' , _A , overwrite_result=re.sub('''DDPM''' , '''Test''' , _A ) , )
71
1
import unittest from transformers import DebertaVaTokenizer, DebertaVaTokenizerFast from transformers.testing_utils import get_tests_dir, require_sentencepiece, require_tokenizers, slow from ...test_tokenization_common import TokenizerTesterMixin _lowerCAmelCase = get_tests_dir("fixtures/spiece.model") @require_sentencepiece @require_tokenizers class lowerCAmelCase_ ( __lowercase, unittest.TestCase ): UpperCAmelCase = DebertaVaTokenizer UpperCAmelCase = DebertaVaTokenizerFast UpperCAmelCase = True UpperCAmelCase = True def UpperCamelCase_ ( self : Optional[int] ): super().setUp() # We have a SentencePiece fixture for testing _UpperCamelCase = DebertaVaTokenizer(_A , unk_token='''<unk>''' ) tokenizer.save_pretrained(self.tmpdirname ) def UpperCamelCase_ ( self : List[Any] , _A : Dict ): _UpperCamelCase = '''this is a test''' _UpperCamelCase = '''this is a test''' return input_text, output_text def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = '''<pad>''' _UpperCamelCase = 0 self.assertEqual(self.get_tokenizer()._convert_token_to_id(_A ) , _A ) self.assertEqual(self.get_tokenizer()._convert_id_to_token(_A ) , _A ) def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = list(self.get_tokenizer().get_vocab().keys() ) self.assertEqual(vocab_keys[0] , '''<pad>''' ) self.assertEqual(vocab_keys[1] , '''<unk>''' ) self.assertEqual(vocab_keys[-1] , '''[PAD]''' ) self.assertEqual(len(_A ) , 3_0001 ) def UpperCamelCase_ ( self : Dict ): self.assertEqual(self.get_tokenizer().vocab_size , 3_0000 ) def UpperCamelCase_ ( self : Optional[Any] ): # fmt: off _UpperCamelCase = ''' \tHeLLo!how \n Are yoU? ''' _UpperCamelCase = ['''▁hello''', '''!''', '''how''', '''▁are''', '''▁you''', '''?'''] # fmt: on _UpperCamelCase = DebertaVaTokenizer(_A , do_lower_case=_A ) _UpperCamelCase = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) _UpperCamelCase = DebertaVaTokenizerFast(_A , do_lower_case=_A ) _UpperCamelCase = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) @unittest.skip('''There is an inconsistency between slow and fast tokenizer due to a bug in the fast one.''' ) def UpperCamelCase_ ( self : Any ): pass @unittest.skip('''There is an inconsistency between slow and fast tokenizer due to a bug in the fast one.''' ) def UpperCamelCase_ ( self : Any ): pass def UpperCamelCase_ ( self : Tuple ): # fmt: off _UpperCamelCase = '''I was born in 92000, and this is falsé.''' _UpperCamelCase = ['''▁''', '''<unk>''', '''▁was''', '''▁born''', '''▁in''', '''▁9''', '''2000''', '''▁''', ''',''', '''▁and''', '''▁this''', '''▁is''', '''▁fal''', '''s''', '''<unk>''', '''▁''', '''.''', ] # fmt: on _UpperCamelCase = DebertaVaTokenizer(_A , split_by_punct=_A ) _UpperCamelCase = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) _UpperCamelCase = DebertaVaTokenizerFast(_A , split_by_punct=_A ) _UpperCamelCase = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : List[str] ): # fmt: off _UpperCamelCase = '''I was born in 92000, and this is falsé.''' _UpperCamelCase = ['''▁i''', '''▁was''', '''▁born''', '''▁in''', '''▁9''', '''2000''', '''▁''', ''',''', '''▁and''', '''▁this''', '''▁is''', '''▁fal''', '''s''', '''<unk>''', '''▁''', '''.''', ] # fmt: on _UpperCamelCase = DebertaVaTokenizer(_A , do_lower_case=_A , split_by_punct=_A ) _UpperCamelCase = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) _UpperCamelCase = DebertaVaTokenizerFast(_A , do_lower_case=_A , split_by_punct=_A ) _UpperCamelCase = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : Optional[Any] ): # fmt: off _UpperCamelCase = '''I was born in 92000, and this is falsé.''' _UpperCamelCase = ['''▁i''', '''▁was''', '''▁born''', '''▁in''', '''▁9''', '''2000''', ''',''', '''▁and''', '''▁this''', '''▁is''', '''▁fal''', '''s''', '''<unk>''', '''.''', ] # fmt: on _UpperCamelCase = DebertaVaTokenizer(_A , do_lower_case=_A , split_by_punct=_A ) _UpperCamelCase = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) _UpperCamelCase = DebertaVaTokenizerFast(_A , do_lower_case=_A , split_by_punct=_A ) _UpperCamelCase = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : int ): # fmt: off _UpperCamelCase = '''I was born in 92000, and this is falsé.''' _UpperCamelCase = ['''▁''', '''<unk>''', '''▁was''', '''▁born''', '''▁in''', '''▁9''', '''2000''', '''▁''', ''',''', '''▁and''', '''▁this''', '''▁is''', '''▁fal''', '''s''', '''<unk>''', '''▁''', '''.''', ] # fmt: on _UpperCamelCase = DebertaVaTokenizer(_A , do_lower_case=_A , split_by_punct=_A ) _UpperCamelCase = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) _UpperCamelCase = DebertaVaTokenizerFast(_A , do_lower_case=_A , split_by_punct=_A ) _UpperCamelCase = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : str ): # fmt: off _UpperCamelCase = ''' \tHeLLo!how \n Are yoU? ''' _UpperCamelCase = ['''▁''', '''<unk>''', '''e''', '''<unk>''', '''o''', '''!''', '''how''', '''▁''', '''<unk>''', '''re''', '''▁yo''', '''<unk>''', '''?'''] # fmt: on _UpperCamelCase = DebertaVaTokenizer(_A , do_lower_case=_A , split_by_punct=_A ) _UpperCamelCase = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) _UpperCamelCase = DebertaVaTokenizerFast(_A , do_lower_case=_A , split_by_punct=_A ) _UpperCamelCase = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = self.get_tokenizer() _UpperCamelCase = self.get_rust_tokenizer() _UpperCamelCase = '''I was born in 92000, and this is falsé.''' _UpperCamelCase = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) ) _UpperCamelCase = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) ) self.assertListEqual(_A , _A ) _UpperCamelCase = tokenizer.encode(_A , add_special_tokens=_A ) _UpperCamelCase = rust_tokenizer.encode(_A , add_special_tokens=_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = self.get_rust_tokenizer() _UpperCamelCase = tokenizer.encode(_A ) _UpperCamelCase = rust_tokenizer.encode(_A ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : str ): _UpperCamelCase = '''This is a test''' _UpperCamelCase = [13, 1, 4398, 25, 21, 1289] _UpperCamelCase = ['''▁''', '''T''', '''his''', '''▁is''', '''▁a''', '''▁test'''] _UpperCamelCase = ['''▁''', '''<unk>''', '''his''', '''▁is''', '''▁a''', '''▁test'''] _UpperCamelCase = DebertaVaTokenizer(_A , keep_accents=_A ) _UpperCamelCase = DebertaVaTokenizerFast(_A , keep_accents=_A ) _UpperCamelCase = tokenizer.encode(_A , add_special_tokens=_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = tokenizer.tokenize(_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = tokenizer.convert_ids_to_tokens(_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = rust_tokenizer.encode(_A , add_special_tokens=_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = rust_tokenizer.tokenize(_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = rust_tokenizer.convert_ids_to_tokens(_A ) self.assertListEqual(_A , _A ) # fmt: off _UpperCamelCase = '''I was born in 92000, and this is falsé.''' _UpperCamelCase = [13, 1, 23, 386, 19, 561, 3050, 15, 17, 48, 25, 8256, 18, 1, 9] _UpperCamelCase = ['''▁''', '''I''', '''▁was''', '''▁born''', '''▁in''', '''▁9''', '''2000''', ''',''', '''▁and''', '''▁this''', '''▁is''', '''▁fal''', '''s''', '''é''', '''.''', ] _UpperCamelCase = ['''▁''', '''<unk>''', '''▁was''', '''▁born''', '''▁in''', '''▁9''', '''2000''', ''',''', '''▁and''', '''▁this''', '''▁is''', '''▁fal''', '''s''', '''<unk>''', '''.''', ] # fmt: on _UpperCamelCase = tokenizer.encode(_A , add_special_tokens=_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = tokenizer.tokenize(_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = tokenizer.convert_ids_to_tokens(_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = rust_tokenizer.encode(_A , add_special_tokens=_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = rust_tokenizer.tokenize(_A ) self.assertListEqual(_A , _A ) _UpperCamelCase = rust_tokenizer.convert_ids_to_tokens(_A ) self.assertListEqual(_A , _A ) def UpperCamelCase_ ( self : List[str] ): _UpperCamelCase = DebertaVaTokenizer(_A ) _UpperCamelCase = tokenizer.encode('''sequence builders''' ) _UpperCamelCase = tokenizer.encode('''multi-sequence build''' ) _UpperCamelCase = tokenizer.build_inputs_with_special_tokens(_A ) _UpperCamelCase = tokenizer.build_inputs_with_special_tokens(_A , _A ) self.assertEqual([tokenizer.cls_token_id] + text + [tokenizer.sep_token_id] , _A ) self.assertEqual( [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id] + text_a + [tokenizer.sep_token_id] , _A , ) @slow def UpperCamelCase_ ( self : List[str] ): # fmt: off _UpperCamelCase = {'''input_ids''': [[1, 3_9867, 36, 1_9390, 486, 27, 3_5052, 8_1436, 18, 6_0685, 1225, 7, 3_5052, 8_1436, 18, 9367, 1_6899, 18, 1_5937, 53, 594, 773, 18, 1_6287, 3_0465, 36, 1_5937, 6, 4_1139, 38, 3_6979, 6_0763, 191, 6, 3_4132, 99, 6, 5_0538, 390, 4_3230, 6, 3_4132, 2779, 2_0850, 14, 699, 1072, 1194, 36, 382, 1_0901, 53, 7, 699, 1072, 2084, 36, 2_0422, 630, 53, 19, 105, 3049, 1896, 1053, 1_6899, 1506, 11, 3_7978, 4243, 7, 1237, 3_1869, 200, 1_6566, 654, 6, 3_5052, 8_1436, 7, 5_5630, 1_3593, 4, 2], [1, 26, 1_5011, 13, 667, 8, 1053, 18, 2_3611, 1237, 7_2356, 1_2820, 34, 10_4134, 1209, 35, 1_3313, 6627, 21, 202, 347, 7, 164, 2399, 11, 46, 4485, 4, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 5, 1232, 2864, 1_5785, 1_4951, 105, 5, 8581, 1250, 4, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], '''token_type_ids''': [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], '''attention_mask''': [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]} # noqa: E501 # fmt: on self.tokenizer_integration_test_util( expected_encoding=_A , model_name='''microsoft/deberta-v2-xlarge''' , revision='''ad6e42c1532ddf3a15c39246b63f5559d558b670''' , )
71
from __future__ import annotations import math class lowerCAmelCase_ : def __init__( self : int , _A : int ): _UpperCamelCase = size # approximate the overall size of segment tree with given value _UpperCamelCase = [0 for i in range(0 , 4 * size )] # create array to store lazy update _UpperCamelCase = [0 for i in range(0 , 4 * size )] _UpperCamelCase = [0 for i in range(0 , 4 * size )] # flag for lazy update def UpperCamelCase_ ( self : str , _A : int ): return idx * 2 def UpperCamelCase_ ( self : Any , _A : int ): return idx * 2 + 1 def UpperCamelCase_ ( self : Union[str, Any] , _A : int , _A : int , _A : int , _A : list[int] ): if left_element == right_element: _UpperCamelCase = a[left_element - 1] else: _UpperCamelCase = (left_element + right_element) // 2 self.build(self.left(_A ) , _A , _A , _A ) self.build(self.right(_A ) , mid + 1 , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) def UpperCamelCase_ ( self : Tuple , _A : int , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return True if left_element >= a and right_element <= b: _UpperCamelCase = val if left_element != right_element: _UpperCamelCase = val _UpperCamelCase = val _UpperCamelCase = True _UpperCamelCase = True return True _UpperCamelCase = (left_element + right_element) // 2 self.update(self.left(_A ) , _A , _A , _A , _A , _A ) self.update(self.right(_A ) , mid + 1 , _A , _A , _A , _A ) _UpperCamelCase = max( self.segment_tree[self.left(_A )] , self.segment_tree[self.right(_A )] ) return True def UpperCamelCase_ ( self : Any , _A : int , _A : int , _A : int , _A : int , _A : int ): if self.flag[idx] is True: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = False if left_element != right_element: _UpperCamelCase = self.lazy[idx] _UpperCamelCase = self.lazy[idx] _UpperCamelCase = True _UpperCamelCase = True if right_element < a or left_element > b: return -math.inf if left_element >= a and right_element <= b: return self.segment_tree[idx] _UpperCamelCase = (left_element + right_element) // 2 _UpperCamelCase = self.query(self.left(_A ) , _A , _A , _A , _A ) _UpperCamelCase = self.query(self.right(_A ) , mid + 1 , _A , _A , _A ) return max(_A , _A ) def __str__( self : Tuple ): return str([self.query(1 , 1 , self.size , _A , _A ) for i in range(1 , self.size + 1 )] ) if __name__ == "__main__": _lowerCAmelCase = [1, 2, -4, 7, 3, -5, 6, 11, -20, 9, 14, 15, 5, 2, -8] _lowerCAmelCase = 15 _lowerCAmelCase = SegmentTree(size) segt.build(1, 1, size, A) print(segt.query(1, 1, size, 4, 6)) print(segt.query(1, 1, size, 7, 11)) print(segt.query(1, 1, size, 7, 12)) segt.update(1, 1, size, 1, 3, 111) print(segt.query(1, 1, size, 1, 15)) segt.update(1, 1, size, 7, 8, 235) print(segt)
71
1
import os from shutil import copyfile from typing import Any, Dict, List, Optional, Tuple import sentencepiece as spm from ...tokenization_utils import PreTrainedTokenizer from ...utils import logging _lowerCAmelCase = logging.get_logger(__name__) _lowerCAmelCase = {"vocab_file": "spiece.model"} _lowerCAmelCase = { "vocab_file": { "bert_for_seq_generation": ( "https://huggingface.co/google/bert_for_seq_generation_L-24_bbc_encoder/resolve/main/spiece.model" ), } } _lowerCAmelCase = {"bert_for_seq_generation": 512} class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = VOCAB_FILES_NAMES UpperCAmelCase = PRETRAINED_VOCAB_FILES_MAP UpperCAmelCase = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES UpperCAmelCase = [] UpperCAmelCase = ["input_ids", "attention_mask"] def __init__( self : Any , _A : List[str] , _A : List[Any]="<s>" , _A : List[str]="</s>" , _A : List[str]="<unk>" , _A : Any="<pad>" , _A : int="<::::>" , _A : Optional[Dict[str, Any]] = None , **_A : List[Any] , ): _UpperCamelCase = {} if sp_model_kwargs is None else sp_model_kwargs # Add extra_ids to the special token list super().__init__( bos_token=_A , eos_token=_A , unk_token=_A , pad_token=_A , sep_token=_A , sp_model_kwargs=self.sp_model_kwargs , **_A , ) _UpperCamelCase = vocab_file _UpperCamelCase = spm.SentencePieceProcessor(**self.sp_model_kwargs ) self.sp_model.Load(_A ) @property def UpperCamelCase_ ( self : Optional[Any] ): return self.sp_model.get_piece_size() def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = {self.convert_ids_to_tokens(_A ): i for i in range(self.vocab_size )} vocab.update(self.added_tokens_encoder ) return vocab def __getstate__( self : Optional[int] ): _UpperCamelCase = self.__dict__.copy() _UpperCamelCase = None return state def __setstate__( self : List[Any] , _A : List[Any] ): _UpperCamelCase = d # for backward compatibility if not hasattr(self , '''sp_model_kwargs''' ): _UpperCamelCase = {} _UpperCamelCase = spm.SentencePieceProcessor(**self.sp_model_kwargs ) self.sp_model.Load(self.vocab_file ) def UpperCamelCase_ ( self : Optional[Any] , _A : str ): return self.sp_model.encode(_A , out_type=_A ) def UpperCamelCase_ ( self : Dict , _A : List[str] ): return self.sp_model.piece_to_id(_A ) def UpperCamelCase_ ( self : Union[str, Any] , _A : Any ): _UpperCamelCase = self.sp_model.IdToPiece(_A ) return token def UpperCamelCase_ ( self : str , _A : Dict ): _UpperCamelCase = [] _UpperCamelCase = '''''' for token in tokens: # make sure that special tokens are not decoded using sentencepiece model if token in self.all_special_tokens: out_string += self.sp_model.decode(_A ) + token _UpperCamelCase = [] else: current_sub_tokens.append(_A ) out_string += self.sp_model.decode(_A ) return out_string.strip() def UpperCamelCase_ ( self : Union[str, Any] , _A : str , _A : Optional[str] = None ): if not os.path.isdir(_A ): logger.error(F"""Vocabulary path ({save_directory}) should be a directory""" ) return _UpperCamelCase = os.path.join( _A , (filename_prefix + '''-''' if filename_prefix else '''''') + VOCAB_FILES_NAMES['''vocab_file'''] ) if os.path.abspath(self.vocab_file ) != os.path.abspath(_A ) and os.path.isfile(self.vocab_file ): copyfile(self.vocab_file , _A ) elif not os.path.isfile(self.vocab_file ): with open(_A , '''wb''' ) as fi: _UpperCamelCase = self.sp_model.serialized_model_proto() fi.write(_A ) return (out_vocab_file,)
71
from typing import TYPE_CHECKING from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available _lowerCAmelCase = { "configuration_jukebox": [ "JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP", "JukeboxConfig", "JukeboxPriorConfig", "JukeboxVQVAEConfig", ], "tokenization_jukebox": ["JukeboxTokenizer"], } try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: _lowerCAmelCase = [ "JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST", "JukeboxModel", "JukeboxPreTrainedModel", "JukeboxVQVAE", "JukeboxPrior", ] if TYPE_CHECKING: from .configuration_jukebox import ( JUKEBOX_PRETRAINED_CONFIG_ARCHIVE_MAP, JukeboxConfig, JukeboxPriorConfig, JukeboxVQVAEConfig, ) from .tokenization_jukebox import JukeboxTokenizer try: if not is_torch_available(): raise OptionalDependencyNotAvailable() except OptionalDependencyNotAvailable: pass else: from .modeling_jukebox import ( JUKEBOX_PRETRAINED_MODEL_ARCHIVE_LIST, JukeboxModel, JukeboxPreTrainedModel, JukeboxPrior, JukeboxVQVAE, ) else: import sys _lowerCAmelCase = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
71
1
from graphs.minimum_spanning_tree_kruskal import kruskal def _snake_case ( ): _UpperCamelCase = 9 _UpperCamelCase = [ [0, 1, 4], [0, 7, 8], [1, 2, 8], [7, 8, 7], [7, 6, 1], [2, 8, 2], [8, 6, 6], [2, 3, 7], [2, 5, 4], [6, 5, 2], [3, 5, 14], [3, 4, 9], [5, 4, 10], [1, 7, 11], ] _UpperCamelCase = kruskal(__snake_case , __snake_case ) _UpperCamelCase = [ [7, 6, 1], [2, 8, 2], [6, 5, 2], [0, 1, 4], [2, 5, 4], [2, 3, 7], [0, 7, 8], [3, 4, 9], ] assert sorted(__snake_case ) == sorted(__snake_case )
71
import multiprocessing import os from typing import BinaryIO, Optional, Union import fsspec from .. import Dataset, Features, NamedSplit, config from ..formatting import query_table from ..packaged_modules.json.json import Json from ..utils import logging from ..utils.typing import NestedDataStructureLike, PathLike from .abc import AbstractDatasetReader class lowerCAmelCase_ ( __lowercase ): def __init__( self : int , _A : NestedDataStructureLike[PathLike] , _A : Optional[NamedSplit] = None , _A : Optional[Features] = None , _A : str = None , _A : bool = False , _A : bool = False , _A : Optional[str] = None , _A : Optional[int] = None , **_A : str , ): super().__init__( _A , split=_A , features=_A , cache_dir=_A , keep_in_memory=_A , streaming=_A , num_proc=_A , **_A , ) _UpperCamelCase = field _UpperCamelCase = path_or_paths if isinstance(_A , _A ) else {self.split: path_or_paths} _UpperCamelCase = Json( cache_dir=_A , data_files=_A , features=_A , field=_A , **_A , ) def UpperCamelCase_ ( self : List[str] ): # Build iterable dataset if self.streaming: _UpperCamelCase = self.builder.as_streaming_dataset(split=self.split ) # Build regular (map-style) dataset else: _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = None self.builder.download_and_prepare( download_config=_A , download_mode=_A , verification_mode=_A , base_path=_A , num_proc=self.num_proc , ) _UpperCamelCase = self.builder.as_dataset( split=self.split , verification_mode=_A , in_memory=self.keep_in_memory ) return dataset class lowerCAmelCase_ : def __init__( self : Optional[Any] , _A : Dataset , _A : Union[PathLike, BinaryIO] , _A : Optional[int] = None , _A : Optional[int] = None , **_A : List[str] , ): if num_proc is not None and num_proc <= 0: raise ValueError(F"""num_proc {num_proc} must be an integer > 0.""" ) _UpperCamelCase = dataset _UpperCamelCase = path_or_buf _UpperCamelCase = batch_size if batch_size else config.DEFAULT_MAX_BATCH_SIZE _UpperCamelCase = num_proc _UpperCamelCase = '''utf-8''' _UpperCamelCase = to_json_kwargs def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = self.to_json_kwargs.pop('''path_or_buf''' , _A ) _UpperCamelCase = self.to_json_kwargs.pop('''orient''' , '''records''' ) _UpperCamelCase = self.to_json_kwargs.pop('''lines''' , True if orient == '''records''' else False ) _UpperCamelCase = self.to_json_kwargs.pop('''index''' , False if orient in ['''split''', '''table'''] else True ) _UpperCamelCase = self.to_json_kwargs.pop('''compression''' , _A ) if compression not in [None, "infer", "gzip", "bz2", "xz"]: raise NotImplementedError(F"""`datasets` currently does not support {compression} compression""" ) if isinstance(self.path_or_buf , (str, bytes, os.PathLike) ): with fsspec.open(self.path_or_buf , '''wb''' , compression=_A ) as buffer: _UpperCamelCase = self._write(file_obj=_A , orient=_A , lines=_A , index=_A , **self.to_json_kwargs ) else: if compression: raise NotImplementedError( F"""The compression parameter is not supported when writing to a buffer, but compression={compression}""" ''' was passed. Please provide a local path instead.''' ) _UpperCamelCase = self._write( file_obj=self.path_or_buf , orient=_A , lines=_A , index=_A , **self.to_json_kwargs ) return written def UpperCamelCase_ ( self : Any , _A : Optional[Any] ): _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = args _UpperCamelCase = query_table( table=self.dataset.data , key=slice(_A , offset + self.batch_size ) , indices=self.dataset._indices , ) _UpperCamelCase = batch.to_pandas().to_json( path_or_buf=_A , orient=_A , lines=_A , index=_A , **_A ) if not json_str.endswith('''\n''' ): json_str += "\n" return json_str.encode(self.encoding ) def UpperCamelCase_ ( self : int , _A : BinaryIO , _A : Dict , _A : Optional[Any] , _A : Dict , **_A : str , ): _UpperCamelCase = 0 if self.num_proc is None or self.num_proc == 1: for offset in logging.tqdm( range(0 , len(self.dataset ) , self.batch_size ) , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating json from Arrow format''' , ): _UpperCamelCase = self._batch_json((offset, orient, lines, index, to_json_kwargs) ) written += file_obj.write(_A ) else: _UpperCamelCase , _UpperCamelCase = len(self.dataset ), self.batch_size with multiprocessing.Pool(self.num_proc ) as pool: for json_str in logging.tqdm( pool.imap( self._batch_json , [(offset, orient, lines, index, to_json_kwargs) for offset in range(0 , _A , _A )] , ) , total=(num_rows // batch_size) + 1 if num_rows % batch_size else num_rows // batch_size , unit='''ba''' , disable=not logging.is_progress_bar_enabled() , desc='''Creating json from Arrow format''' , ): written += file_obj.write(_A ) return written
71
1
from __future__ import annotations import inspect import unittest import numpy as np from transformers import DeiTConfig from transformers.testing_utils import require_tf, require_vision, slow from transformers.utils import cached_property, is_tf_available, is_vision_available from ...test_configuration_common import ConfigTester from ...test_modeling_tf_common import TFModelTesterMixin, floats_tensor, ids_tensor from ...test_pipeline_mixin import PipelineTesterMixin if is_tf_available(): import tensorflow as tf from transformers import ( TFDeiTForImageClassification, TFDeiTForImageClassificationWithTeacher, TFDeiTForMaskedImageModeling, TFDeiTModel, ) from transformers.models.deit.modeling_tf_deit import TF_DEIT_PRETRAINED_MODEL_ARCHIVE_LIST if is_vision_available(): from PIL import Image from transformers import DeiTImageProcessor class lowerCAmelCase_ : def __init__( self : List[str] , _A : List[Any] , _A : List[Any]=13 , _A : List[Any]=30 , _A : Union[str, Any]=2 , _A : List[Any]=3 , _A : List[Any]=True , _A : Any=True , _A : Dict=32 , _A : Optional[Any]=2 , _A : Optional[Any]=4 , _A : List[str]=37 , _A : Union[str, Any]="gelu" , _A : List[Any]=0.1 , _A : Union[str, Any]=0.1 , _A : Optional[int]=10 , _A : int=0.02 , _A : List[str]=3 , _A : Tuple=None , _A : Any=2 , ): _UpperCamelCase = parent _UpperCamelCase = batch_size _UpperCamelCase = image_size _UpperCamelCase = patch_size _UpperCamelCase = num_channels _UpperCamelCase = is_training _UpperCamelCase = use_labels _UpperCamelCase = hidden_size _UpperCamelCase = num_hidden_layers _UpperCamelCase = num_attention_heads _UpperCamelCase = intermediate_size _UpperCamelCase = hidden_act _UpperCamelCase = hidden_dropout_prob _UpperCamelCase = attention_probs_dropout_prob _UpperCamelCase = type_sequence_label_size _UpperCamelCase = initializer_range _UpperCamelCase = scope _UpperCamelCase = encoder_stride # in DeiT, the seq length equals the number of patches + 2 (we add 2 for the [CLS] and distilation tokens) _UpperCamelCase = (image_size // patch_size) ** 2 _UpperCamelCase = num_patches + 2 def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] ) _UpperCamelCase = None if self.use_labels: _UpperCamelCase = ids_tensor([self.batch_size] , self.type_sequence_label_size ) _UpperCamelCase = self.get_config() return config, pixel_values, labels def UpperCamelCase_ ( self : List[Any] ): return DeiTConfig( image_size=self.image_size , patch_size=self.patch_size , num_channels=self.num_channels , hidden_size=self.hidden_size , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , is_decoder=_A , initializer_range=self.initializer_range , encoder_stride=self.encoder_stride , ) def UpperCamelCase_ ( self : Tuple , _A : Tuple , _A : Dict , _A : Dict ): _UpperCamelCase = TFDeiTModel(config=_A ) _UpperCamelCase = model(_A ) self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) ) def UpperCamelCase_ ( self : str , _A : int , _A : List[str] , _A : int ): _UpperCamelCase = TFDeiTForMaskedImageModeling(config=_A ) _UpperCamelCase = model(_A ) self.parent.assertEqual( result.reconstruction.shape , (self.batch_size, self.num_channels, self.image_size, self.image_size) ) # test greyscale images _UpperCamelCase = 1 _UpperCamelCase = TFDeiTForMaskedImageModeling(_A ) _UpperCamelCase = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] ) _UpperCamelCase = model(_A ) self.parent.assertEqual(result.reconstruction.shape , (self.batch_size, 1, self.image_size, self.image_size) ) def UpperCamelCase_ ( self : Union[str, Any] , _A : Optional[Any] , _A : Any , _A : Tuple ): _UpperCamelCase = self.type_sequence_label_size _UpperCamelCase = TFDeiTForImageClassification(_A ) _UpperCamelCase = model(_A , labels=_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) ) # test greyscale images _UpperCamelCase = 1 _UpperCamelCase = TFDeiTForImageClassification(_A ) _UpperCamelCase = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] ) _UpperCamelCase = model(_A , labels=_A ) self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) ) def UpperCamelCase_ ( self : str ): _UpperCamelCase = self.prepare_config_and_inputs() _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = config_and_inputs _UpperCamelCase = {'''pixel_values''': pixel_values} return config, inputs_dict @require_tf class lowerCAmelCase_ ( __lowercase, __lowercase, unittest.TestCase ): UpperCAmelCase = ( ( TFDeiTModel, TFDeiTForImageClassification, TFDeiTForImageClassificationWithTeacher, TFDeiTForMaskedImageModeling, ) if is_tf_available() else () ) UpperCAmelCase = ( { "feature-extraction": TFDeiTModel, "image-classification": (TFDeiTForImageClassification, TFDeiTForImageClassificationWithTeacher), } if is_tf_available() else {} ) UpperCAmelCase = False UpperCAmelCase = False UpperCAmelCase = False UpperCAmelCase = False def UpperCamelCase_ ( self : Optional[Any] ): _UpperCamelCase = TFDeiTModelTester(self ) _UpperCamelCase = ConfigTester(self , config_class=_A , has_text_modality=_A , hidden_size=37 ) def UpperCamelCase_ ( self : int ): self.config_tester.run_common_tests() @unittest.skip(reason='''DeiT does not use inputs_embeds''' ) def UpperCamelCase_ ( self : Union[str, Any] ): pass def UpperCamelCase_ ( self : Tuple ): _UpperCamelCase , _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() for model_class in self.all_model_classes: _UpperCamelCase = model_class(_A ) self.assertIsInstance(model.get_input_embeddings() , (tf.keras.layers.Layer) ) _UpperCamelCase = model.get_output_embeddings() self.assertTrue(x is None or isinstance(_A , tf.keras.layers.Dense ) ) def UpperCamelCase_ ( self : Optional[int] ): _UpperCamelCase , _UpperCamelCase = self.model_tester.prepare_config_and_inputs_for_common() for model_class in self.all_model_classes: _UpperCamelCase = model_class(_A ) _UpperCamelCase = inspect.signature(model.call ) # signature.parameters is an OrderedDict => so arg_names order is deterministic _UpperCamelCase = [*signature.parameters.keys()] _UpperCamelCase = ['''pixel_values'''] self.assertListEqual(arg_names[:1] , _A ) def UpperCamelCase_ ( self : Union[str, Any] ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model(*_A ) def UpperCamelCase_ ( self : int ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_masked_image_modeling(*_A ) def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_image_classification(*_A ) def UpperCamelCase_ ( self : Any , _A : Union[str, Any] , _A : List[Any] , _A : Any=False ): _UpperCamelCase = super()._prepare_for_class(_A , _A , return_labels=_A ) if return_labels: if "labels" in inputs_dict and "labels" not in inspect.signature(model_class.call ).parameters: del inputs_dict["labels"] return inputs_dict @slow def UpperCamelCase_ ( self : int ): for model_name in TF_DEIT_PRETRAINED_MODEL_ARCHIVE_LIST[:1]: _UpperCamelCase = TFDeiTModel.from_pretrained(_A ) self.assertIsNotNone(_A ) def _snake_case ( ): _UpperCamelCase = Image.open('''./tests/fixtures/tests_samples/COCO/000000039769.png''' ) return image @require_tf @require_vision class lowerCAmelCase_ ( unittest.TestCase ): @cached_property def UpperCamelCase_ ( self : List[str] ): return ( DeiTImageProcessor.from_pretrained('''facebook/deit-base-distilled-patch16-224''' ) if is_vision_available() else None ) @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = TFDeiTForImageClassificationWithTeacher.from_pretrained('''facebook/deit-base-distilled-patch16-224''' ) _UpperCamelCase = self.default_image_processor _UpperCamelCase = prepare_img() _UpperCamelCase = image_processor(images=_A , return_tensors='''tf''' ) # forward pass _UpperCamelCase = model(**_A ) # verify the logits _UpperCamelCase = tf.TensorShape((1, 1000) ) self.assertEqual(outputs.logits.shape , _A ) _UpperCamelCase = tf.constant([-1.0266, 0.1912, -1.2861] ) self.assertTrue(np.allclose(outputs.logits[0, :3] , _A , atol=1e-4 ) )
71
import enum import warnings from .. import MODEL_FOR_CAUSAL_LM_MAPPING, TF_MODEL_FOR_CAUSAL_LM_MAPPING from ..utils import add_end_docstrings, is_tf_available from .base import PIPELINE_INIT_ARGS, Pipeline if is_tf_available(): import tensorflow as tf class lowerCAmelCase_ ( enum.Enum ): UpperCAmelCase = 0 UpperCAmelCase = 1 UpperCAmelCase = 2 @add_end_docstrings(__lowercase ) class lowerCAmelCase_ ( __lowercase ): UpperCAmelCase = "\n In 1991, the remains of Russian Tsar Nicholas II and his family (except for Alexei and Maria) are discovered. The\n voice of Nicholas's young son, Tsarevich Alexei Nikolaevich, narrates the remainder of the story. 1883 Western\n Siberia, a young Grigori Rasputin is asked by his father and a group of men to perform magic. Rasputin has a vision\n and denounces one of the men as a horse thief. Although his father initially slaps him for making such an\n accusation, Rasputin watches as the man is chased outside and beaten. Twenty years later, Rasputin sees a vision of\n the Virgin Mary, prompting him to become a priest. Rasputin quickly becomes famous, with people, even a bishop,\n begging for his blessing. <eod> </s> <eos>\n " def __init__( self : Tuple , *_A : List[str] , **_A : str ): super().__init__(*_A , **_A ) self.check_model_type( TF_MODEL_FOR_CAUSAL_LM_MAPPING if self.framework == '''tf''' else MODEL_FOR_CAUSAL_LM_MAPPING ) if "prefix" not in self._preprocess_params: # This is very specific. The logic is quite complex and needs to be done # as a "default". # It also defines both some preprocess_kwargs and generate_kwargs # which is why we cannot put them in their respective methods. _UpperCamelCase = None if self.model.config.prefix is not None: _UpperCamelCase = self.model.config.prefix if prefix is None and self.model.__class__.__name__ in [ "XLNetLMHeadModel", "TransfoXLLMHeadModel", "TFXLNetLMHeadModel", "TFTransfoXLLMHeadModel", ]: # For XLNet and TransformerXL we add an article to the prompt to give more state to the model. _UpperCamelCase = self.XL_PREFIX if prefix is not None: # Recalculate some generate_kwargs linked to prefix. _UpperCamelCase , _UpperCamelCase , _UpperCamelCase = self._sanitize_parameters(prefix=_A , **self._forward_params ) _UpperCamelCase = {**self._preprocess_params, **preprocess_params} _UpperCamelCase = {**self._forward_params, **forward_params} def UpperCamelCase_ ( self : Dict , _A : Optional[int]=None , _A : Any=None , _A : Optional[int]=None , _A : List[str]=None , _A : List[Any]=None , _A : int=None , _A : Tuple=None , _A : Optional[Any]=None , **_A : Optional[int] , ): _UpperCamelCase = {} if prefix is not None: _UpperCamelCase = prefix if prefix: _UpperCamelCase = self.tokenizer( _A , padding=_A , add_special_tokens=_A , return_tensors=self.framework ) _UpperCamelCase = prefix_inputs['''input_ids'''].shape[-1] if handle_long_generation is not None: if handle_long_generation not in {"hole"}: raise ValueError( F"""{handle_long_generation} is not a valid value for `handle_long_generation` parameter expected""" ''' [None, \'hole\']''' ) _UpperCamelCase = handle_long_generation preprocess_params.update(_A ) _UpperCamelCase = generate_kwargs _UpperCamelCase = {} if return_full_text is not None and return_type is None: if return_text is not None: raise ValueError('''`return_text` is mutually exclusive with `return_full_text`''' ) if return_tensors is not None: raise ValueError('''`return_full_text` is mutually exclusive with `return_tensors`''' ) _UpperCamelCase = ReturnType.FULL_TEXT if return_full_text else ReturnType.NEW_TEXT if return_tensors is not None and return_type is None: if return_text is not None: raise ValueError('''`return_text` is mutually exclusive with `return_tensors`''' ) _UpperCamelCase = ReturnType.TENSORS if return_type is not None: _UpperCamelCase = return_type if clean_up_tokenization_spaces is not None: _UpperCamelCase = clean_up_tokenization_spaces if stop_sequence is not None: _UpperCamelCase = self.tokenizer.encode(_A , add_special_tokens=_A ) if len(_A ) > 1: warnings.warn( '''Stopping on a multiple token sequence is not yet supported on transformers. The first token of''' ''' the stop sequence will be used as the stop sequence string in the interim.''' ) _UpperCamelCase = stop_sequence_ids[0] return preprocess_params, forward_params, postprocess_params def UpperCamelCase_ ( self : int , *_A : Union[str, Any] , **_A : Union[str, Any] ): # Parse arguments if self.model.__class__.__name__ in ["TransfoXLLMHeadModel"]: kwargs.update({'''add_space_before_punct_symbol''': True} ) return super()._parse_and_tokenize(*_A , **_A ) def __call__( self : List[str] , _A : str , **_A : Any ): return super().__call__(_A , **_A ) def UpperCamelCase_ ( self : Optional[Any] , _A : List[str] , _A : int="" , _A : Optional[Any]=None , **_A : Optional[Any] ): _UpperCamelCase = self.tokenizer( prefix + prompt_text , padding=_A , add_special_tokens=_A , return_tensors=self.framework ) _UpperCamelCase = prompt_text if handle_long_generation == "hole": _UpperCamelCase = inputs['''input_ids'''].shape[-1] if "max_new_tokens" in generate_kwargs: _UpperCamelCase = generate_kwargs['''max_new_tokens'''] else: _UpperCamelCase = generate_kwargs.get('''max_length''' , self.model.config.max_length ) - cur_len if new_tokens < 0: raise ValueError('''We cannot infer how many new tokens are expected''' ) if cur_len + new_tokens > self.tokenizer.model_max_length: _UpperCamelCase = self.tokenizer.model_max_length - new_tokens if keep_length <= 0: raise ValueError( '''We cannot use `hole` to handle this generation the number of desired tokens exceeds the''' ''' models max length''' ) _UpperCamelCase = inputs['''input_ids'''][:, -keep_length:] if "attention_mask" in inputs: _UpperCamelCase = inputs['''attention_mask'''][:, -keep_length:] return inputs def UpperCamelCase_ ( self : Dict , _A : Optional[int] , **_A : str ): _UpperCamelCase = model_inputs['''input_ids'''] _UpperCamelCase = model_inputs.get('''attention_mask''' , _A ) # Allow empty prompts if input_ids.shape[1] == 0: _UpperCamelCase = None _UpperCamelCase = None _UpperCamelCase = 1 else: _UpperCamelCase = input_ids.shape[0] _UpperCamelCase = model_inputs.pop('''prompt_text''' ) # If there is a prefix, we may need to adjust the generation length. Do so without permanently modifying # generate_kwargs, as some of the parameterization may come from the initialization of the pipeline. _UpperCamelCase = generate_kwargs.pop('''prefix_length''' , 0 ) if prefix_length > 0: _UpperCamelCase = '''max_new_tokens''' in generate_kwargs or ( '''generation_config''' in generate_kwargs and generate_kwargs['''generation_config'''].max_new_tokens is not None ) if not has_max_new_tokens: _UpperCamelCase = generate_kwargs.get('''max_length''' ) or self.model.config.max_length generate_kwargs["max_length"] += prefix_length _UpperCamelCase = '''min_new_tokens''' in generate_kwargs or ( '''generation_config''' in generate_kwargs and generate_kwargs['''generation_config'''].min_new_tokens is not None ) if not has_min_new_tokens and "min_length" in generate_kwargs: generate_kwargs["min_length"] += prefix_length # BS x SL _UpperCamelCase = self.model.generate(input_ids=_A , attention_mask=_A , **_A ) _UpperCamelCase = generated_sequence.shape[0] if self.framework == "pt": _UpperCamelCase = generated_sequence.reshape(_A , out_b // in_b , *generated_sequence.shape[1:] ) elif self.framework == "tf": _UpperCamelCase = tf.reshape(_A , (in_b, out_b // in_b, *generated_sequence.shape[1:]) ) return {"generated_sequence": generated_sequence, "input_ids": input_ids, "prompt_text": prompt_text} def UpperCamelCase_ ( self : List[str] , _A : Dict , _A : Optional[Any]=ReturnType.FULL_TEXT , _A : Dict=True ): _UpperCamelCase = model_outputs['''generated_sequence'''][0] _UpperCamelCase = model_outputs['''input_ids'''] _UpperCamelCase = model_outputs['''prompt_text'''] _UpperCamelCase = generated_sequence.numpy().tolist() _UpperCamelCase = [] for sequence in generated_sequence: if return_type == ReturnType.TENSORS: _UpperCamelCase = {'''generated_token_ids''': sequence} elif return_type in {ReturnType.NEW_TEXT, ReturnType.FULL_TEXT}: # Decode text _UpperCamelCase = self.tokenizer.decode( _A , skip_special_tokens=_A , clean_up_tokenization_spaces=_A , ) # Remove PADDING prompt of the sequence if XLNet or Transfo-XL model is used if input_ids is None: _UpperCamelCase = 0 else: _UpperCamelCase = len( self.tokenizer.decode( input_ids[0] , skip_special_tokens=_A , clean_up_tokenization_spaces=_A , ) ) if return_type == ReturnType.FULL_TEXT: _UpperCamelCase = prompt_text + text[prompt_length:] else: _UpperCamelCase = text[prompt_length:] _UpperCamelCase = {'''generated_text''': all_text} records.append(_A ) return records
71
1
import unittest from transformers import is_torch_available, is_vision_available from transformers.testing_utils import require_torch, require_vision, slow, torch_device if is_torch_available(): import torch from transformers import AutoModelForImageClassification if is_vision_available(): from transformers import AutoImageProcessor @require_torch @require_vision class lowerCAmelCase_ ( unittest.TestCase ): @slow def UpperCamelCase_ ( self : Dict ): _UpperCamelCase = AutoImageProcessor.from_pretrained('''microsoft/dit-base-finetuned-rvlcdip''' ) _UpperCamelCase = AutoModelForImageClassification.from_pretrained('''microsoft/dit-base-finetuned-rvlcdip''' ) model.to(_A ) from datasets import load_dataset _UpperCamelCase = load_dataset('''nielsr/rvlcdip-demo''' ) _UpperCamelCase = dataset['''train'''][0]['''image'''].convert('''RGB''' ) _UpperCamelCase = image_processor(_A , return_tensors='''pt''' ).to(_A ) # forward pass with torch.no_grad(): _UpperCamelCase = model(**_A ) _UpperCamelCase = outputs.logits _UpperCamelCase = torch.Size((1, 16) ) self.assertEqual(logits.shape , _A ) _UpperCamelCase = torch.tensor( [-0.4158, -0.4092, -0.4347] , device=_A , dtype=torch.float , ) self.assertTrue(torch.allclose(logits[0, :3] , _A , atol=1e-4 ) )
71
import unittest from transformers import is_torch_available from transformers.testing_utils import require_sentencepiece, require_tokenizers, require_torch, slow, torch_device if is_torch_available(): from transformers import AutoModelForSeqaSeqLM, AutoTokenizer @require_torch @require_sentencepiece @require_tokenizers class lowerCAmelCase_ ( unittest.TestCase ): @slow def UpperCamelCase_ ( self : Any ): _UpperCamelCase = AutoModelForSeqaSeqLM.from_pretrained('''google/mt5-small''' , return_dict=_A ).to(_A ) _UpperCamelCase = AutoTokenizer.from_pretrained('''google/mt5-small''' ) _UpperCamelCase = tokenizer('''Hello there''' , return_tensors='''pt''' ).input_ids _UpperCamelCase = tokenizer('''Hi I am''' , return_tensors='''pt''' ).input_ids _UpperCamelCase = model(input_ids.to(_A ) , labels=labels.to(_A ) ).loss _UpperCamelCase = -(labels.shape[-1] * loss.item()) _UpperCamelCase = -84.9127 self.assertTrue(abs(mtf_score - EXPECTED_SCORE ) < 1e-4 )
71
1