Upload ParchmentForCausalLM

README.md

@@ -0,0 +1,199 @@
+---
+library_name: transformers
+tags: []
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]

config.json

@@ -0,0 +1,27 @@
+{
+  "architectures": [
+    "ParchmentForCausalLM"
+  ],
+  "auto_map": {
+    "AutoConfig": "configuration_parchment.ParchmentConfig",
+    "AutoModelForCausalLM": "modeling_parchment.ParchmentForCausalLM"
+  },
+  "bos_token_id": 100257,
+  "d_ff": 3072,
+  "d_model": 768,
+  "dtype": "bfloat16",
+  "eos_token_id": 100257,
+  "hidden_size": 768,
+  "max_seq_len": 1024,
+  "model_type": "parchment",
+  "n_heads": 12,
+  "n_layers": 12,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 100276,
+  "rms_norm_eps": 1e-06,
+  "rope_base": 10000.0,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.8.1",
+  "vocab_size": 100277
+}

configuration_parchment.py

@@ -0,0 +1,30 @@
+from transformers import PretrainedConfig
+
+
+class ParchmentConfig(PretrainedConfig):
+    model_type = "parchment"
+
+    def __init__(
+        self,
+        vocab_size: int = 100277,
+        d_model: int = 768,
+        n_heads: int = 12,
+        n_layers: int = 12,
+        max_seq_len: int = 1024,
+        rms_norm_eps: float = 1e-6,
+        rope_base: float = 10000.0,
+        tie_word_embeddings: bool = True,
+        **kwargs,
+    ):
+        self.vocab_size = vocab_size
+        self.d_model = d_model
+        self.n_heads = n_heads
+        self.n_layers = n_layers
+        self.max_seq_len = max_seq_len
+        self.rms_norm_eps = rms_norm_eps
+        self.rope_base = rope_base
+        # aliases expected by transformers internals
+        self.num_hidden_layers = n_layers
+        self.hidden_size = d_model
+        self.num_attention_heads = n_heads
+        super().__init__(tie_word_embeddings=tie_word_embeddings, **kwargs)

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 100257,
+  "eos_token_id": 100257,
+  "output_attentions": false,
+  "output_hidden_states": false,
+  "pad_token_id": 100257,
+  "transformers_version": "5.8.1"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8f0365945a84407219115b99318c2a65cd90e494e0334673d0c77d5a7b6c3f7
+size 323947000

modeling_parchment.py

@@ -0,0 +1,202 @@
+import math
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from transformers import PreTrainedModel
+from transformers.generation import GenerationMixin
+from transformers.modeling_outputs import CausalLMOutputWithPast
+
+from .configuration_parchment import ParchmentConfig
+
+
+class Embeddings(nn.Module):
+    def __init__(self, vocab_size, d_model):
+        super().__init__()
+        self.embeds = nn.Embedding(vocab_size, d_model)
+        nn.init.normal_(self.embeds.weight, mean=0, std=d_model ** -0.5)
+
+    def forward(self, token_ids):
+        return self.embeds(token_ids)
+
+
+class RoPE(nn.Module):
+    def __init__(self, d_k, max_seq_len, base=10000.0):
+        super().__init__()
+        self.max_seq_len = max_seq_len
+        # persistent=True so inv_freq is saved/loaded via from_pretrained
+        inv_freq = 1.0 / (base ** (torch.arange(0, d_k, 2).float() / d_k))
+        self.register_buffer("inv_freq", inv_freq, persistent=True)
+        self._cos_cache: torch.Tensor | None = None
+        self._sin_cache: torch.Tensor | None = None
+
+    def _build_cache(self, device: torch.device, dtype: torch.dtype):
+        t = torch.arange(self.max_seq_len, device=device, dtype=torch.float32)
+        freqs = torch.outer(t, self.inv_freq.to(device, torch.float32))
+        emb = torch.cat([freqs, freqs], dim=-1)
+        self._cos_cache = emb.cos()[None, None].to(dtype)
+        self._sin_cache = emb.sin()[None, None].to(dtype)
+
+    def rotate_half(self, x):
+        x1, x2 = x[..., : x.shape[-1] // 2], x[..., x.shape[-1] // 2 :]
+        return torch.cat([-x2, x1], dim=-1)
+
+    def forward(self, q, k):
+        if self._cos_cache is None or self._cos_cache.device != q.device or self._cos_cache.dtype != q.dtype:
+            self._build_cache(q.device, q.dtype)
+        seq = q.shape[2]
+        cos = self._cos_cache[:, :, :seq]
+        sin = self._sin_cache[:, :, :seq]
+        q = (q * cos) + (self.rotate_half(q) * sin)
+        k = (k * cos) + (self.rotate_half(k) * sin)
+        return q, k
+
+
+class RMSNorm(nn.Module):
+    def __init__(self, d_model, eps=1e-6):
+        super().__init__()
+        self.scale = nn.Parameter(torch.ones(d_model))
+        self.eps = eps
+
+    def forward(self, x):
+        return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps) * self.scale
+
+
+class MultiHeadAttention(nn.Module):
+    def __init__(self, n_heads, d_model, max_seq_len, rope_base):
+        super().__init__()
+        self.n_heads = n_heads
+        self.d_k = d_model // n_heads
+        self.d_model = d_model
+        self.W_Q = nn.Linear(d_model, d_model, bias=False)
+        self.W_K = nn.Linear(d_model, d_model, bias=False)
+        self.W_V = nn.Linear(d_model, d_model, bias=False)
+        self.W_O = nn.Linear(d_model, d_model, bias=False)
+        self.rope = RoPE(self.d_k, max_seq_len, base=rope_base)
+        self.W_O.RESIDUAL_SCALE_INIT = True
+
+    def forward(self, x):
+        B, T, _ = x.shape
+        Q = self.W_Q(x).view(B, T, self.n_heads, self.d_k).permute(0, 2, 1, 3)
+        K = self.W_K(x).view(B, T, self.n_heads, self.d_k).permute(0, 2, 1, 3)
+        V = self.W_V(x).view(B, T, self.n_heads, self.d_k).permute(0, 2, 1, 3)
+        Q, K = self.rope(Q, K)
+        out = F.scaled_dot_product_attention(Q, K, V, is_causal=True)
+        out = out.permute(0, 2, 1, 3).contiguous().view(B, T, self.d_model)
+        return self.W_O(out)
+
+
+class SwiGLU(nn.Module):
+    def __init__(self, d_model):
+        super().__init__()
+        hidden = int(2 / 3 * 4 * d_model)
+        hidden = (hidden + 63) // 64 * 64
+        self.w1 = nn.Linear(d_model, hidden, bias=False)
+        self.w2 = nn.Linear(hidden, d_model, bias=False)
+        self.w3 = nn.Linear(d_model, hidden, bias=False)
+        self.w2.RESIDUAL_SCALE_INIT = True
+
+    def forward(self, x):
+        return self.w2(F.silu(self.w1(x)) * self.w3(x))
+
+
+class TransformerBlock(nn.Module):
+    def __init__(self, d_model, n_heads, max_seq_len, rope_base):
+        super().__init__()
+        self.attn = MultiHeadAttention(n_heads, d_model, max_seq_len, rope_base)
+        self.ff = SwiGLU(d_model)
+        self.norm1 = RMSNorm(d_model)
+        self.norm2 = RMSNorm(d_model)
+
+    def forward(self, x):
+        x = x + self.attn(self.norm1(x))
+        x = x + self.ff(self.norm2(x))
+        return x
+
+
+class ParchmentModel(PreTrainedModel):
+    config_class = ParchmentConfig
+    base_model_prefix = "model"
+
+    def __init__(self, config: ParchmentConfig):
+        super().__init__(config)
+        self.embeddings = Embeddings(config.vocab_size, config.d_model)
+        self.blocks = nn.ModuleList([
+            TransformerBlock(config.d_model, config.n_heads, config.max_seq_len, config.rope_base)
+            for _ in range(config.n_layers)
+        ])
+        self.norm = RMSNorm(config.d_model, eps=config.rms_norm_eps)
+        self.post_init()
+
+    def _init_weights(self, module):
+        if isinstance(module, nn.Linear):
+            std = 0.02
+            if hasattr(module, "RESIDUAL_SCALE_INIT"):
+                std /= math.sqrt(2 * self.config.n_layers)
+            nn.init.normal_(module.weight, mean=0.0, std=std)
+            if module.bias is not None:
+                nn.init.zeros_(module.bias)
+        elif isinstance(module, nn.Embedding):
+            nn.init.normal_(module.weight, mean=0, std=self.config.d_model ** -0.5)
+
+    def forward(self, input_ids: torch.LongTensor) -> torch.Tensor:
+        x = self.embeddings(input_ids)
+        for block in self.blocks:
+            x = block(x)
+        return self.norm(x)
+
+
+class ParchmentForCausalLM(PreTrainedModel, GenerationMixin):
+    config_class = ParchmentConfig
+    base_model_prefix = "model"
+    _tied_weights_keys = {"lm_head.weight": "model.embeddings.embeds.weight"}
+    _keys_to_ignore_on_load_missing = [r"lm_head\.weight", r".*\.rope\."]
+    _supports_cache_class = False
+    _supports_static_cache = False
+
+    def __init__(self, config: ParchmentConfig):
+        super().__init__(config)
+        self.model = ParchmentModel(config)
+        self.lm_head = nn.Linear(config.d_model, config.vocab_size, bias=False)
+        self.post_init()
+
+    def get_input_embeddings(self):
+        return self.model.embeddings.embeds
+
+    def set_input_embeddings(self, value):
+        self.model.embeddings.embeds = value
+
+    def get_output_embeddings(self):
+        return self.lm_head
+
+    def set_output_embeddings(self, value):
+        self.lm_head = value
+
+    def _init_weights(self, module):
+        self.model._init_weights(module)
+
+    def forward(
+        self,
+        input_ids: torch.LongTensor,
+        attention_mask: torch.Tensor | None = None,
+        labels: torch.LongTensor | None = None,
+        **kwargs,
+    ) -> CausalLMOutputWithPast:
+        hidden = self.model(input_ids)
+        logits = self.lm_head(hidden)
+
+        loss = None
+        if labels is not None:
+            shift_logits = logits[:, :-1, :].contiguous()
+            shift_labels = labels[:, 1:].contiguous()
+            loss = F.cross_entropy(
+                shift_logits.view(-1, self.config.vocab_size),
+                shift_labels.view(-1),
+            )
+
+        return CausalLMOutputWithPast(loss=loss, logits=logits)
+
+    def prepare_inputs_for_generation(self, input_ids, attention_mask=None, **kwargs):
+        return {
+            "input_ids": input_ids[:, -self.config.max_seq_len:],
+            "attention_mask": attention_mask,
+        }