File size: 8,623 Bytes
b754bbe |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 |
from dataclasses import dataclass, field
from typing import Optional
from result_parser import yes_or_no, find_option_number, anomaly_detection, trajectory_prediction, trajectory_classification
result_parsers = {
"poi_category_recognition": find_option_number,
"poi_identification": yes_or_no,
"urban_region_function_recognition": find_option_number,
"administrative_region_determination": find_option_number,
"point_trajectory": find_option_number,
"point_region": find_option_number,
"trajectory_region": find_option_number,
"trajectory_identification": yes_or_no,
"trajectory_trajectory": find_option_number,
"direction_determination": find_option_number,
"trajectory_anomaly_detection": anomaly_detection,
"trajectory_classification": trajectory_classification,
"trajectory_prediction": trajectory_prediction
}
max_tokens = {
"poi_category_recognition": 15,
"poi_identification": 15,
"urban_region_function_recognition": 15,
"administrative_region_determination": 15,
"point_trajectory": 15,
"point_region": 15,
"trajectory_region": 15,
"trajectory_identification": 15,
"trajectory_trajectory": 15,
"direction_determination": 15,
"trajectory_anomaly_detection": 15,
"trajectory_classification": 15,
"trajectory_prediction": 50
}
dataset_files = {
"poi_category_recognition": ["../datasets/basic/knowledge_comprehension/poi_category_recognition.jsonl"],
"poi_identification": ["../datasets/basic/knowledge_comprehension/poi_identification.jsonl"],
"urban_region_function_recognition": ["../datasets/basic/knowledge_comprehension/urban_region_function_recognition.jsonl"],
"administrative_region_determination": ["../datasets/basic/knowledge_comprehension/administrative_region_determination.jsonl"],
"point_trajectory": ["../datasets/basic/spatiotemporal_reasoning/point_trajectory.jsonl"],
"point_region": ["../datasets/basic/spatiotemporal_reasoning/point_region_2regions.jsonl",
"../datasets/basic/spatiotemporal_reasoning/point_region_3regions.jsonl",
"../datasets/basic/spatiotemporal_reasoning/point_region_4regions.jsonl",
"../datasets/basic/spatiotemporal_reasoning/point_region_5regions.jsonl"],
"trajectory_region": ["../datasets/basic/spatiotemporal_reasoning/trajectory_region_length2.jsonl",
"../datasets/basic/spatiotemporal_reasoning/trajectory_region_length4.jsonl",
"../datasets/basic/spatiotemporal_reasoning/trajectory_region_length6.jsonl",
"../datasets/basic/spatiotemporal_reasoning/trajectory_region_length8.jsonl",
"../datasets/basic/spatiotemporal_reasoning/trajectory_region_length10.jsonl"],
"trajectory_identification": ["../datasets/basic/spatiotemporal_reasoning/trajectory_identification_downsampling.jsonl",
"../datasets/basic/spatiotemporal_reasoning/trajectory_identification_staggered_sampling.jsonl",
"../datasets/basic/spatiotemporal_reasoning/trajectory_identification_spatial_offset.jsonl",
"../datasets/basic/spatiotemporal_reasoning/trajectory_identification_temporal_offset.jsonl"],
"trajectory_trajectory": ["../datasets/basic/accurate_calculation/trajectory_trajectory.jsonl"],
"direction_determination": ["../datasets/basic/accurate_calculation/direction_determination.jsonl"],
"trajectory_anomaly_detection": ["../datasets/basic/downstream_applications/trajectory_anomaly_detection_abnormal.jsonl",
"../datasets/basic/downstream_applications/trajectory_anomaly_detection_normal.jsonl"],
"trajectory_classification": ["../datasets/basic/downstream_applications/trajectory_classification.jsonl"],
"trajectory_prediction": ["../datasets/basic/downstream_applications/trajectory_prediction.jsonl"]
}
icl_files = {
"poi_identification": "../datasets/icl/poi_identification.jsonl",
"trajectory_region": "../datasets/icl/trajectory_region.jsonl",
"trajectory_trajectory": "../datasets/icl/trajectory_trajectory.jsonl",
"direction_determination": "../datasets/icl/direction_determination.jsonl",
"trajectory_anomaly_detection": "../datasets/icl/trajectory_anomaly_detection.jsonl",
"trajectory_prediction": "../datasets/icl/trajectory_prediction.jsonl"
}
cot_files = {
"urban_region_function_recognition": "../datasets/cot/urban_region_function_recognition.jsonl",
"trajectory_region": "../datasets/cot/trajectory_region.jsonl",
"trajectory_trajectory": "../datasets/cot/trajectory_trajectory.jsonl",
"trajectory_classification": "../datasets/cot/trajectory_classification.jsonl"
}
sft_files = {
"administrative_region_determination": {
"train": "../datasets/sft/administrative_region_determination_train.jsonl",
"valid": "../datasets/sft/administrative_region_determination_valid.jsonl"
},
"direction_determination": {
"train": "../datasets/sft/direction_determination_train.jsonl",
"valid": "../datasets/sft/direction_determination_valid.jsonl"
},
"trajectory_anomaly_detection":{
"train": "../datasets/sft/trajectory_anomaly_detection_train.jsonl",
"valid": "../datasets/sft/trajectory_anomaly_detection_valid.jsonl"
},
"trajectory_prediction": {
"train": "../datasets/sft/trajectory_prediction_train.jsonl",
"valid": "../datasets/sft/trajectory_prediction_valid.jsonl"
},
"trajectory_region": {
"train": "../datasets/sft/trajectory_region_train.jsonl",
"valid": "../datasets/sft/trajectory_region_valid.jsonl"
},
"trajectory_trajectory": {
"train": "../datasets/sft/trajectory_trajectory_train.jsonl",
"valid": "../datasets/sft/trajectory_trajectory_valid.jsonl"
}
}
@dataclass
class ScriptArguments:
"""
These arguments vary depending on how many GPUs you have, what their capacity and features are, and what size model you want to train.
"""
per_device_train_batch_size: Optional[int] = field(default=4)
per_device_eval_batch_size: Optional[int] = field(default=1)
gradient_accumulation_steps: Optional[int] = field(default=4)
learning_rate: Optional[float] = field(default=2e-4)
max_grad_norm: Optional[float] = field(default=0.3)
weight_decay: Optional[int] = field(default=0.001)
lora_alpha: Optional[int] = field(default=16)
lora_dropout: Optional[float] = field(default=0.1)
lora_r: Optional[int] = field(default=8)
max_seq_length: Optional[int] = field(default=2048)
model_name: Optional[str] = field(
default=None,
metadata={
"help": "The model that you want to train from the Hugging Face hub. E.g. gpt2, gpt2-xl, bert, etc."
}
)
dataset_name: Optional[str] = field(
default="stingning/ultrachat",
metadata={"help": "The preference dataset to use."},
)
fp16: Optional[bool] = field(
default=False,
metadata={"help": "Enables fp16 training."},
)
bf16: Optional[bool] = field(
default=False,
metadata={"help": "Enables bf16 training."},
)
packing: Optional[bool] = field(
default=True,
metadata={"help": "Use packing dataset creating."},
)
gradient_checkpointing: Optional[bool] = field(
default=True,
metadata={"help": "Enables gradient checkpointing."},
)
use_flash_attention_2: Optional[bool] = field(
default=False,
metadata={"help": "Enables Flash Attention 2."},
)
optim: Optional[str] = field(
default="paged_adamw_32bit",
metadata={"help": "The optimizer to use."},
)
lr_scheduler_type: str = field(
default="constant",
metadata={"help": "Learning rate schedule. Constant a bit better than cosine, and has advantage for analysis"},
)
max_steps: int = field(default=1000, metadata={"help": "How many optimizer update steps to take"})
warmup_ratio: float = field(default=0.03, metadata={"help": "Fraction of steps to do a warmup for"})
save_steps: int = field(default=100, metadata={"help": "Save checkpoint every X updates steps."})
logging_steps: int = field(default=10, metadata={"help": "Log every X updates steps."})
output_dir: str = field(
default="./results",
metadata={"help": "The output directory where the model predictions and checkpoints will be written."},
)
|