File size: 11,580 Bytes
1a7ee60 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 | """
Training Script for Document Classifier
Fine-tunes MiniLM on CORD, SROIE, and FUNSD datasets
Includes data augmentation and early stopping
"""
import json
import logging
import os
from typing import Dict, List
import numpy as np
import torch
import torch.nn as nn
from sklearn.metrics import accuracy_score, classification_report, f1_score
from sklearn.model_selection import train_test_split
from torch.optim import AdamW
from torch.utils.data import DataLoader, Dataset
from tqdm import tqdm
from transformers import AutoTokenizer, get_linear_schedule_with_warmup
from classifier_model import DocumentClassifier
from dataset_loader import UnifiedDatasetLoader
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
class DocumentClassificationDataset(Dataset):
"""PyTorch Dataset for document classification"""
def __init__(
self, texts: List[str], labels: List[int], tokenizer, max_length: int = 512
):
self.texts = texts
self.labels = labels
self.tokenizer = tokenizer
self.max_length = max_length
def __len__(self):
return len(self.texts)
def __getitem__(self, idx):
text = self.texts[idx]
label = self.labels[idx]
# Tokenize
encoding = self.tokenizer(
text,
max_length=self.max_length,
padding="max_length",
truncation=True,
return_tensors="pt",
)
return {
"input_ids": encoding["input_ids"].flatten(),
"attention_mask": encoding["attention_mask"].flatten(),
"labels": torch.tensor(label, dtype=torch.long),
}
class TextAugmenter:
"""Simple text augmentation for document classification"""
@staticmethod
def random_token_masking(text: str, mask_prob: float = 0.1) -> str:
"""Randomly mask tokens with [MASK]"""
tokens = text.split()
num_to_mask = int(len(tokens) * mask_prob)
if num_to_mask > 0:
mask_indices = np.random.choice(len(tokens), num_to_mask, replace=False)
for idx in mask_indices:
tokens[idx] = "[MASK]"
return " ".join(tokens)
@staticmethod
def random_deletion(text: str, delete_prob: float = 0.1) -> str:
"""Randomly delete tokens"""
tokens = text.split()
tokens = [t for t in tokens if np.random.random() > delete_prob]
return " ".join(tokens) if tokens else text
class ClassifierTrainer:
"""Trainer for document classifier"""
def __init__(
self,
model_name: str = "nreimers/MiniLM-L6-H384-uncased",
output_dir: str = "models/classifier",
device: str = None,
):
self.model_name = model_name
self.output_dir = output_dir
self.device = (
device if device else ("cuda" if torch.cuda.is_available() else "cpu")
)
os.makedirs(output_dir, exist_ok=True)
logger.info(f"Trainer initialized on device: {self.device}")
def prepare_data(
self, augment: bool = True, test_size: float = 0.1, val_size: float = 0.1
):
"""Load and prepare datasets"""
logger.info("Loading datasets...")
# Load data from all sources
loader = UnifiedDatasetLoader()
train_data = loader.load_classification_dataset(
datasets=["cord", "sroie", "funsd"], split="train"
)
# Get label mappings
mappings = loader.get_label_mappings()
self.label2id = mappings["classification"]
self.id2label = mappings["classification_id2label"]
# Extract texts and labels
texts = [item["text"] for item in train_data]
labels = [self.label2id[item["label"]] for item in train_data]
# Apply augmentation
if augment:
logger.info("Applying data augmentation...")
augmenter = TextAugmenter()
aug_texts = []
aug_labels = []
for text, label in zip(texts, labels):
# Original
aug_texts.append(text)
aug_labels.append(label)
# Augmented versions (30% of data)
if np.random.random() < 0.3:
aug_texts.append(augmenter.random_token_masking(text))
aug_labels.append(label)
if np.random.random() < 0.3:
aug_texts.append(augmenter.random_deletion(text))
aug_labels.append(label)
texts = aug_texts
labels = aug_labels
logger.info(f"Augmented dataset size: {len(texts)}")
# Split into train/val/test
train_texts, temp_texts, train_labels, temp_labels = train_test_split(
texts,
labels,
test_size=(test_size + val_size),
random_state=42,
stratify=labels,
)
val_texts, test_texts, val_labels, test_labels = train_test_split(
temp_texts,
temp_labels,
test_size=test_size / (test_size + val_size),
random_state=42,
stratify=temp_labels,
)
logger.info(
f"Train: {len(train_texts)}, Val: {len(val_texts)}, Test: {len(test_texts)}"
)
return (
(train_texts, train_labels),
(val_texts, val_labels),
(test_texts, test_labels),
)
def train(
self,
train_data: tuple,
val_data: tuple,
num_epochs: int = 10,
batch_size: int = 16,
learning_rate: float = 2e-5,
warmup_steps: int = 500,
early_stopping_patience: int = 3,
):
"""Train the classifier"""
train_texts, train_labels = train_data
val_texts, val_labels = val_data
# Load tokenizer and model
tokenizer = AutoTokenizer.from_pretrained(self.model_name)
model = DocumentClassifier(
model_name=self.model_name, num_labels=len(self.label2id)
)
model.to(self.device)
# Create datasets
train_dataset = DocumentClassificationDataset(
train_texts, train_labels, tokenizer
)
val_dataset = DocumentClassificationDataset(val_texts, val_labels, tokenizer)
# Create dataloaders
train_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True)
val_loader = DataLoader(val_dataset, batch_size=batch_size)
# Optimizer and scheduler
optimizer = AdamW(model.parameters(), lr=learning_rate)
total_steps = len(train_loader) * num_epochs
scheduler = get_linear_schedule_with_warmup(
optimizer, num_warmup_steps=warmup_steps, num_training_steps=total_steps
)
# Training loop
best_val_accuracy = 0
patience_counter = 0
for epoch in range(num_epochs):
logger.info(f"\nEpoch {epoch + 1}/{num_epochs}")
# Training
model.train()
train_loss = 0
train_preds = []
train_true = []
progress_bar = tqdm(train_loader, desc="Training")
for batch in progress_bar:
optimizer.zero_grad()
input_ids = batch["input_ids"].to(self.device)
attention_mask = batch["attention_mask"].to(self.device)
labels = batch["labels"].to(self.device)
outputs = model(input_ids, attention_mask, labels)
loss = outputs["loss"]
loss.backward()
optimizer.step()
scheduler.step()
train_loss += loss.item()
preds = torch.argmax(outputs["logits"], dim=-1)
train_preds.extend(preds.cpu().numpy())
train_true.extend(labels.cpu().numpy())
progress_bar.set_postfix({"loss": loss.item()})
avg_train_loss = train_loss / len(train_loader)
train_accuracy = accuracy_score(train_true, train_preds)
train_f1 = f1_score(train_true, train_preds, average="weighted")
logger.info(
f"Train Loss: {avg_train_loss:.4f}, Accuracy: {train_accuracy:.4f}, F1: {train_f1:.4f}"
)
# Validation
model.eval()
val_loss = 0
val_preds = []
val_true = []
with torch.no_grad():
for batch in tqdm(val_loader, desc="Validation"):
input_ids = batch["input_ids"].to(self.device)
attention_mask = batch["attention_mask"].to(self.device)
labels = batch["labels"].to(self.device)
outputs = model(input_ids, attention_mask, labels)
val_loss += outputs["loss"].item()
preds = torch.argmax(outputs["logits"], dim=-1)
val_preds.extend(preds.cpu().numpy())
val_true.extend(labels.cpu().numpy())
avg_val_loss = val_loss / len(val_loader)
val_accuracy = accuracy_score(val_true, val_preds)
val_f1 = f1_score(val_true, val_preds, average="weighted")
logger.info(
f"Val Loss: {avg_val_loss:.4f}, Accuracy: {val_accuracy:.4f}, F1: {val_f1:.4f}"
)
# Early stopping
if val_accuracy > best_val_accuracy:
best_val_accuracy = val_accuracy
patience_counter = 0
# Save best model
model_path = os.path.join(self.output_dir, "best_classifier.pt")
torch.save(model.state_dict(), model_path)
logger.info(f"Saved best model with accuracy: {best_val_accuracy:.4f}")
else:
patience_counter += 1
if patience_counter >= early_stopping_patience:
logger.info(f"Early stopping triggered after {epoch + 1} epochs")
break
# Save final model and metadata
torch.save(
model.state_dict(), os.path.join(self.output_dir, "final_classifier.pt")
)
metadata = {
"model_name": self.model_name,
"num_labels": len(self.label2id),
"label2id": self.label2id,
"id2label": self.id2label,
"best_val_accuracy": best_val_accuracy,
}
with open(os.path.join(self.output_dir, "metadata.json"), "w") as f:
json.dump(metadata, f, indent=2)
logger.info("Training complete!")
return best_val_accuracy
if __name__ == "__main__":
# Training configuration
trainer = ClassifierTrainer(
model_name="nreimers/MiniLM-L6-H384-uncased", output_dir="models/classifier"
)
# Prepare data
train_data, val_data, test_data = trainer.prepare_data(
augment=True, test_size=0.1, val_size=0.1
)
# Train model
best_accuracy = trainer.train(
train_data=train_data,
val_data=val_data,
num_epochs=15,
batch_size=16,
learning_rate=2e-5,
early_stopping_patience=3,
)
print(f"\nBest validation accuracy: {best_accuracy:.4f}")
|