File size: 14,748 Bytes
67068c5 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 | {
"cells": [
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import re\n",
"import torch\n",
"from transformers import AutoTokenizer, AutoModelForTokenClassification\n",
"\n",
"MODEL_NAME = \"/mnt/jeff/InCar/bert/ModernBERT-base\" \n",
"tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)\n",
"\n",
"# 定義新的 NER 標籤類別\n",
"# O: 無關\n",
"# B-NUM: 數字實體的開頭\n",
"# I-NUM: 數字實體的內部\n",
"# B-SWITCH: 開關指令的開頭\n",
"# I-SWITCH: 開關指令的內部\n",
"# B-LEVEL: 最大/最小指令的開頭\n",
"# I-LEVEL: 最大/最小指令的內部\n",
"# B-AREA_ID: 新增的區域識別符號開頭\n",
"# I-AREA_ID: 新增的區域識別符號內部\n",
"labels = [\"O\", \"B-NUM\", \"I-NUM\", \"B-SWITCH\", \"I-SWITCH\", \"B-LEVEL\", \"I-LEVEL\", \"B-AREA_ID\", \"I-AREA_ID\"]\n",
"id2label = {i: label for i, label in enumerate(labels)}\n",
"label_map = {label: i for i, label in enumerate(labels)}\n",
"# 定義關鍵字映射\n",
"# 這裡我們用簡單的字典來模擬模型預測的邏輯\n",
"# 在實際應用中，這些會是從訓練好的模型中得出的結果\n",
"switch_keywords = ['開啟', '開', '打開', '關閉', '關', '關掉']\n",
"level_keywords = ['最大', '最小', '全滿', '全開']\n",
"# 數字正則表達式，用於匹配連續的數字\n",
"number_pattern = r'\\d+'\n",
"\n",
"# 新增: area_id 關鍵字\n",
"area_id_keywords = [\n",
" \"主駕\", \"駕駛座\", \"主駕駛座\", \"第一排左邊\", \"第一排左座\", \"第一排左側\",\n",
" \"副駕\", \"副駕駛座\", \"第一排右座\", \"第一排右邊\", \"第一排右側\",\n",
" \"第二排左側\", \"第二排左座\", \"中間這排左邊\", \"第二排左邊\",\n",
" \"第二排右側\", \"第二排右座\", \"中間這排右邊\", \"第二排右邊\",\n",
" \"第三排左側\", \"第三排左座\", \"最後面這排左邊\", \"最後排左邊\", \"第三排左邊\",\n",
" \"第三排右側\", \"第三排右座\", \"最後面這排右邊\", \"最後排右邊\", \"第三排右邊\",\n",
" \"全車\", \"整台車\", \"整部車\",\n",
" \"第一排\", \"最前面那排\",\n",
" \"第二排\", \"中間那排\",\n",
" \"第三排\", \"最後面那排\",'前','後'\n",
"]\n",
"\n",
"def predict_ner_labels(text):\n",
"\n",
" tokens = [c for c in text]\n",
" \n",
" tags = [\"O\"] * len(tokens)\n",
" \n",
" for keyword in switch_keywords:\n",
" for match in re.finditer(re.escape(keyword), text):\n",
" start, end = match.span()\n",
" tags[start] = 'B-SWITCH'\n",
" for i in range(start + 1, end):\n",
" tags[i] = 'I-SWITCH'\n",
" \n",
" for keyword in level_keywords:\n",
" for match in re.finditer(re.escape(keyword), text):\n",
" start, end = match.span()\n",
" # 標註開頭為 B-LEVEL，其餘為 I-LEVEL\n",
" tags[start] = 'B-LEVEL'\n",
" for i in range(start + 1, end):\n",
" tags[i] = 'I-LEVEL'\n",
" \n",
" # 處理數字\n",
" for match in re.finditer(number_pattern, text):\n",
" start, end = match.span()\n",
" # 標註開頭為 B-NUM，其餘為 I-NUM\n",
" tags[start] = 'B-NUM'\n",
" for i in range(start + 1, end):\n",
" tags[i] = 'I-NUM'\n",
"\n",
" # 新增: 處理 area_id 關鍵字\n",
" for keyword in area_id_keywords:\n",
" for match in re.finditer(re.escape(keyword), text):\n",
" start, end = match.span()\n",
" tags[start] = 'B-AREA_ID'\n",
" for i in range(start + 1, end):\n",
" tags[i] = 'I-AREA_ID'\n",
" \n",
" return tokens, tags"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {},
"outputs": [],
"source": [
"words = []\n",
"tags = []\n",
"data = json.load(open('classifier_data.jsonl'))\n",
"for d in data:\n",
" word,tag = predict_ner_labels(d['text'])\n",
" words.append(word)\n",
" tags.append(tag)"
]
},
{
"cell_type": "code",
"execution_count": 37,
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "533d111e85a1485ea274f241d19f3ae0",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
"Map: 0%| | 0/23005 [00:00<?, ? examples/s]"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "f585919412f344df9a95cc6825539762",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
"Map: 0%| | 0/2557 [00:00<?, ? examples/s]"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Some weights of ModernBertForTokenClassification were not initialized from the model checkpoint at /mnt/jeff/InCar/bert/ModernBERT-base and are newly initialized: ['classifier.bias', 'classifier.weight']\n",
"You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n",
"Detected kernel version 4.18.0, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher.\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"開始訓練模型...\n"
]
},
{
"data": {
"text/html": [
"\n",
" <div>\n",
" \n",
" <progress value='2876' max='2876' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
" [2876/2876 05:27, Epoch 1/1]\n",
" </div>\n",
" <table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: left;\">\n",
" <th>Step</th>\n",
" <th>Training Loss</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>500</td>\n",
" <td>0.071200</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1000</td>\n",
" <td>0.001200</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1500</td>\n",
" <td>0.000400</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2000</td>\n",
" <td>0.000100</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2500</td>\n",
" <td>0.000100</td>\n",
" </tr>\n",
" </tbody>\n",
"</table><p>"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"模型訓練完成！\n",
"正在測試集上進行最終評估...\n"
]
},
{
"data": {
"text/html": [
"\n",
" <div>\n",
" \n",
" <progress value='320' max='320' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
" [320/320 00:11]\n",
" </div>\n",
" "
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"評估結果：\n",
"{'eval_loss': 6.9491370595642366e-06, 'eval_precision': 1.0, 'eval_recall': 1.0, 'eval_f1': 1.0, 'eval_accuracy': 1.0, 'eval_runtime': 11.9921, 'eval_samples_per_second': 213.223, 'eval_steps_per_second': 26.684, 'epoch': 1.0}\n"
]
}
],
"source": [
"import torch\n",
"from datasets import Dataset\n",
"from transformers import AutoTokenizer, AutoModelForTokenClassification, TrainingArguments, Trainer, DataCollatorForTokenClassification\n",
"from evaluate import load\n",
"import numpy as np\n",
"\n",
"MODEL_NAME = \"/mnt/jeff/InCar/bert/ModernBERT-base\" \n",
"\n",
"\n",
"dataset = Dataset.from_dict({\"tokens\": words, \"ner_tags\": [[label_map[tag] for tag in t] for t in tags]})\n",
"\n",
"split_datasets = dataset.train_test_split(test_size=0.1)\n",
"train_dataset = split_datasets['train']\n",
"eval_dataset = split_datasets['test']\n",
"\n",
"# 初始化分詞器\n",
"tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)\n",
"\n",
"# 定義一個函數來處理分詞和標籤對齊\n",
"def tokenize_and_align_labels(examples):\n",
" tokenized_inputs = tokenizer(examples[\"tokens\"], truncation=True, is_split_into_words=True)\n",
" aligned_labels = []\n",
"\n",
" for i, label in enumerate(examples[\"ner_tags\"]):\n",
" word_ids = tokenized_inputs.word_ids(batch_index=i)\n",
" previous_word_idx = None\n",
" label_ids = []\n",
" for word_idx in word_ids:\n",
" if word_idx is None:\n",
" label_ids.append(-100)\n",
" elif word_idx != previous_word_idx:\n",
" label_ids.append(label[word_idx])\n",
" else:\n",
" label_ids.append(-100)\n",
" previous_word_idx = word_idx\n",
" aligned_labels.append(label_ids)\n",
"\n",
" tokenized_inputs[\"labels\"] = aligned_labels\n",
" return tokenized_inputs\n",
"\n",
"# 將處理函數應用於資料集\n",
"tokenized_train_dataset = train_dataset.map(tokenize_and_align_labels, batched=True)\n",
"tokenized_eval_dataset = eval_dataset.map(tokenize_and_align_labels, batched=True)\n",
"data_collator = DataCollatorForTokenClassification(tokenizer=tokenizer)\n",
"\n",
"# 載入預訓練模型\n",
"model = AutoModelForTokenClassification.from_pretrained(\n",
" MODEL_NAME, \n",
" num_labels=len(labels), \n",
" id2label=id2label, \n",
" label2id=label_map\n",
")\n",
"# 定義評估指標\n",
"metric = load(\"seqeval\")\n",
"def compute_metrics(p):\n",
" predictions, labels = p\n",
" predictions = np.argmax(predictions, axis=2)\n",
"\n",
" true_predictions = [\n",
" [id2label[p] for (p, l) in zip(prediction, label) if l != -100]\n",
" for prediction, label in zip(predictions, labels)\n",
" ]\n",
" true_labels = [\n",
" [id2label[l] for (p, l) in zip(prediction, label) if l != -100]\n",
" for prediction, label in zip(predictions, labels)\n",
" ]\n",
"\n",
" results = metric.compute(predictions=true_predictions, references=true_labels)\n",
" return {\n",
" \"precision\": results[\"overall_precision\"],\n",
" \"recall\": results[\"overall_recall\"],\n",
" \"f1\": results[\"overall_f1\"],\n",
" \"accuracy\": results[\"overall_accuracy\"],\n",
" }\n",
"# 設置訓練參數\n",
"training_args = TrainingArguments(\n",
" output_dir=\"./results_ner\", # 訓練結果儲存目錄\n",
" learning_rate=2e-5, # 學習率\n",
" per_device_train_batch_size=8, # 訓練批次大小\n",
" per_device_eval_batch_size=8, # 評估批次大小\n",
" num_train_epochs=1, # 訓練 epochs 數\n",
" weight_decay=0.01, # 權重衰減\n",
" save_strategy=\"epoch\", # 每個 epoch 結束後儲存檢查點\n",
" report_to='none',\n",
" )\n",
"\n",
"# 創建 Trainer 實例\n",
"trainer = Trainer(\n",
" model=model,\n",
" args=training_args,\n",
" train_dataset=tokenized_train_dataset,\n",
" eval_dataset=tokenized_eval_dataset, # 加入評估資料集\n",
" compute_metrics=compute_metrics,\n",
" data_collator=data_collator\n",
")\n",
"\n",
"# 開始訓練\n",
"print(\"開始訓練模型...\")\n",
"trainer.train()\n",
"print(\"模型訓練完成！\")\n",
"\n",
"# 進行最終評估\n",
"print(\"正在測試集上進行最終評估...\")\n",
"evaluation_results = trainer.evaluate()\n",
"print(\"評估結果：\")\n",
"print(evaluation_results)\n",
"\n",
"trainer.save_model(\"./results_ner\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"for d in tokenized_eval_dataset:\n",
" if len(d['tokens'])==len(d['ner_tags'])==len(d['attention_mask'])==:\n",
" print(d)"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"32"
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"len(d['input_ids'])"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"tokens Its len is 9\n",
"ner_tags Its len is 9\n",
"input_ids Its len is 18\n",
"attention_mask Its len is 18\n",
"labels Its len is 18\n"
]
}
],
"source": [
"for k in tokenized_eval_dataset[0]:\n",
" print(k,'Its len is',len(tokenized_eval_dataset[0][k]))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "base",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.12.7"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
|