.4021378118068288:98296403e0cd6dedb7b420b80d0fe80b_69e2ce7c42c82e665fe03e0c.69e2ce9842c82e665fe03e10.69e2ce988b8bd167e2faa239:Trae CN.T(2026/4/18 08:21:44)

.dockerignore

@@ -0,0 +1,63 @@
+.git
+.github
+.ai
+.agents
+.claude
+__pycache__
+*.pyc
+*.pyo
+*.pyd
+.Python
+env
+build
+develop-eggs
+dist
+downloads
+eggs
+.eggs
+lib
+lib64
+parts
+sdist
+var
+wheels
+*.egg-info
+.installed.cfg
+*.egg
+MANIFEST
+.mypy_cache
+.pytest_cache
+.tox
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+.hypothesis
+.idea
+.vscode
+*.swp
+*.swo
+*~
+docs
+notebooks
+tests
+examples
+benchmark
+benchmark_v2
+utils
+i18n
+scripts
+docker
+*.md
+!README.md
+CITATION.cff
+CODE_OF_CONDUCT.md
+CONTRIBUTING.md
+ISSUES.md
+LICENSE
+SECURITY.md
+Makefile
+conftest.py
+doctest_list.txt

Dockerfile

@@ -0,0 +1,56 @@
+FROM python:3.10-slim
+
+LABEL maintainer="Hugging Face Team"
+LABEL description="Docker image for Hugging Face Transformers - Initial Runtime Environment"
+LABEL repository="https://github.com/huggingface/transformers"
+
+ENV PYTHONDONTWRITEBYTECODE=1
+ENV PYTHONUNBUFFERED=1
+ENV PYTHONPATH=/app/src
+ENV UV_PYTHON=/usr/local/bin/python
+ENV HF_HOME=/app/.cache/huggingface
+ENV HF_HUB_OFFLINE=0
+
+WORKDIR /app
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    git \
+    make \
+    build-essential \
+    curl \
+    ca-certificates \
+    pkg-config \
+    && apt-get clean \
+    && rm -rf /var/lib/apt/lists/*
+
+RUN pip install --no-cache-dir -U pip setuptools wheel uv
+
+COPY setup.py README.md ./
+
+RUN uv pip install --no-cache-dir --system \
+    "torch>=2.4" \
+    "huggingface-hub>=1.5.0,<2.0" \
+    "numpy>=1.17" \
+    "packaging>=20.0" \
+    "pyyaml>=5.1" \
+    "regex>=2025.10.22" \
+    "tokenizers>=0.22.0,<=0.23.0" \
+    "safetensors>=0.4.3" \
+    "tqdm>=4.27" \
+    "typer" \
+    "accelerate>=1.1.0"
+
+COPY pyproject.toml ./
+COPY src/ ./src/
+
+RUN uv pip install --no-cache-dir --system .
+
+RUN mkdir -p /app/.cache/huggingface \
+    && chmod -R 777 /app/.cache
+
+RUN python -c "import transformers; print(f'Transformers v{transformers.__version__} installed successfully!')"
+
+EXPOSE 8000
+
+CMD ["python", "-c", "import transformers; print(f'Welcome to Transformers v{transformers.__version__}!')"]
+

examples/pytorch/text-classification/run_classification.py

@@ -408,8 +408,41 @@ def main():
 
     if data_args.label_column_name is not None and data_args.label_column_name != "label":
         for key in raw_datasets:
+            if data_args.label_column_name not in raw_datasets[key].features:
+                raise ValueError(
+                    f"Label column '{data_args.label_column_name}' not found in dataset split '{key}'. "
+                    f"Available columns are: {list(raw_datasets[key].features.keys())}. "
+                    f"Please check your --label_column_name argument."
+                )
             raw_datasets[key] = raw_datasets[key].rename_column(data_args.label_column_name, "label")
 
+    text_columns = ["sentence"] if data_args.text_column_names is None else data_args.text_column_names.split(",")
+    for split in raw_datasets:
+        available_columns = list(raw_datasets[split].features.keys())
+        for col in text_columns:
+            if col not in available_columns:
+                raise ValueError(
+                    f"Text column '{col}' not found in dataset split '{split}'. "
+                    f"Available columns are: {available_columns}. "
+                    f"Please use --text_column_names to specify correct text column name(s)."
+                )
+        if split != "test" and "label" not in available_columns:
+            raise ValueError(
+                f"Label column 'label' not found in dataset split '{split}'. "
+                f"Available columns are: {available_columns}. "
+                f"Please use --label_column_name to specify correct label column name."
+            )
+
+    if not training_args.do_train and not is_regression:
+        if not (hasattr(model.config, "label2id") and model.config.label2id):
+            raise ValueError(
+                "Classification prediction/evaluation without training requires "
+                "the loaded model to have label2id configuration. "
+                "Please use a model fine-tuned on your classification task."
+            )
+        configured_labels = set(model.config.label2id.keys())
+        logger.info(f"Model configured labels: {sorted(configured_labels)}")
+
     # Trying to have good defaults here, don't hesitate to tweak to your needs.
 
     is_regression = (
@@ -543,26 +576,54 @@ def main():
     max_seq_length = min(data_args.max_seq_length, tokenizer.model_max_length)
 
     def multi_labels_to_ids(labels: list[str]) -> list[float]:
-        ids = [0.0] * len(label_to_id)  # BCELoss requires float as target type
+        ids = [0.0] * len(label_to_id)
         for label in labels:
-            ids[label_to_id[label]] = 1.0
+            label_str = str(label)
+            if label_str not in label_to_id:
+                raise ValueError(
+                    f"Label '{label}' not found in label mapping. "
+                    f"Available labels: {list(label_to_id.keys())}. "
+                    f"Please ensure all labels in the dataset match the labels used during training."
+                )
+            ids[label_to_id[label_str]] = 1.0
         return ids
 
     def preprocess_function(examples):
         if data_args.text_column_names is not None:
             text_column_names = data_args.text_column_names.split(",")
-            # join together text columns into "sentence" column
             examples["sentence"] = examples[text_column_names[0]]
             for column in text_column_names[1:]:
                 for i in range(len(examples[column])):
                     examples["sentence"][i] += data_args.text_column_delimiter + examples[column][i]
-        # Tokenize the texts
+
+        if "sentence" not in examples:
+            raise ValueError(
+                "Text column 'sentence' not found in dataset. "
+                "Please use --text_column_names to specify the correct text column name(s)."
+            )
+
+        examples["sentence"] = [str(s) if s is not None else "" for s in examples["sentence"]]
+
         result = tokenizer(examples["sentence"], padding=padding, max_length=max_seq_length, truncation=True)
+
         if label_to_id is not None and "label" in examples:
             if is_multi_label:
                 result["label"] = [multi_labels_to_ids(l) for l in examples["label"]]
             else:
-                result["label"] = [(label_to_id[str(l)] if l != -1 else -1) for l in examples["label"]]
+                processed_labels = []
+                for l in examples["label"]:
+                    if l == -1:
+                        processed_labels.append(-1)
+                        continue
+                    label_str = str(l)
+                    if label_str not in label_to_id:
+                        raise ValueError(
+                            f"Label '{l}' found in dataset is not recognized. "
+                            f"Valid labels are: {sorted(label_to_id.keys())}. "
+                            f"Please ensure all labels match the training label set."
+                        )
+                    processed_labels.append(label_to_id[label_str])
+                result["label"] = processed_labels
         return result
 
     # Running the preprocessing pipeline on all the datasets
@@ -721,10 +782,10 @@ def compute_metrics(p: EvalPrediction):
                         writer.write(f"{index}\t{item:3.3f}\n")
                     elif is_multi_label:
                         # recover from multi-hot encoding
-                        item = [label_list[i] for i in range(len(item)) if item[i] == 1]
+                        item = [model.config.id2label[i] for i in range(len(item)) if item[i] == 1]
                         writer.write(f"{index}\t{item}\n")
                     else:
-                        item = label_list[item]
+                        item = model.config.id2label[item]
                         writer.write(f"{index}\t{item}\n")
         logger.info(f"Predict results saved at {output_predict_file}")
     kwargs = {"finetuned_from": model_args.model_name_or_path, "tasks": "text-classification"}

examples/pytorch/text-classification/test_label_mapping/test.csv

@@ -0,0 +1,7 @@
+text,label
+"This is a bad movie.",negative
+"I really enjoyed this film.",positive
+"Horrible, don't watch.",negative
+"Amazing movie, I recommend it!",positive
+"Terrible acting.",negative
+"Great story!",positive

examples/pytorch/text-classification/test_label_mapping/test_edge_cases.py

@@ -0,0 +1,153 @@
+#!/usr/bin/env python
+"""
+边界情况测试：验证异常处理和报错机制
+"""
+import os
+import sys
+import subprocess
+import tempfile
+import csv
+
+def run_test(args, description, expect_success=True):
+    """运行测试并检查结果"""
+    print(f"\n{'='*60}")
+    print(f"测试: {description}")
+    print(f"{'='*60}")
+
+    cmd = f"cd /Users/pkcha/transformers && python examples/pytorch/text-classification/run_classification.py {args}"
+    print(f"命令: python run_classification.py {args[:100]}...")
+
+    result = subprocess.run(cmd, shell=True, capture_output=True, text=True)
+
+    if result.returncode == 0:
+        print(f"✅ 执行成功 (返回码: 0)")
+        if not expect_success:
+            print(f"⚠️  警告: 预期失败但实际成功!")
+            return False
+        return True
+    else:
+        print(f"❌ 执行失败 (返回码: {result.returncode})")
+        last_lines = result.stderr.strip().split('\n')[-10:]
+        print("\n错误信息（最后10行）:")
+        print("-" * 40)
+        for line in last_lines:
+            if 'ValueError' in line or 'Error' in line or 'error' in line or 'Found' in line:
+                print(f"  >>> {line}")
+            elif line.strip():
+                print(f"  {line}")
+        print("-" * 40)
+
+        if expect_success:
+            print(f"⚠️  警告: 预期成功但实际失败!")
+            return False
+        return True
+
+def create_test_csv(filename, rows):
+    """创建测试CSV文件"""
+    filepath = os.path.join('/Users/pkcha/transformers/examples/pytorch/text-classification/test_label_mapping', filename)
+    with open(filepath, 'w', newline='') as f:
+        writer = csv.writer(f)
+        writer.writerows(rows)
+    return filepath
+
+def main():
+    test_dir = '/Users/pkcha/transformers/examples/pytorch/text-classification/test_label_mapping'
+    os.chdir(test_dir)
+
+    print("开始边界情况测试...")
+
+    # 1. 测试未知标签报错
+    print("\n\n【测试1】未知标签报错")
+    train_csv_bad_label = create_test_csv('train_bad_label.csv', [
+        ['text', 'label'],
+        ['Good movie', 'positive'],
+        ['Bad film', 'unknown_label']
+    ])
+    val_csv = create_test_csv('val_simple.csv', [
+        ['text', 'label'],
+        ['Great!', 'positive'],
+    ])
+
+    args = f"""--model_name_or_path distilbert-base-uncased \
+        --train_file {train_csv_bad_label} \
+        --validation_file {val_csv} \
+        --do_train \
+        --per_device_train_batch_size 2 \
+        --num_train_epochs 1 \
+        --max_seq_length 32 \
+        --output_dir /tmp/test_output 2>&1
+    """
+    run_test(args, "数据包含未知标签，应抛出明确错误", expect_success=False)
+
+    # 2. 测试文本列缺失报错
+    print("\n\n【测试2】文本列缺失报错")
+    train_csv_no_text = create_test_csv('train_no_text.csv', [
+        ['wrong_column', 'label'],
+        ['Good movie', 'positive'],
+        ['Bad film', 'negative']
+    ])
+
+    args = f"""--model_name_or_path distilbert-base-uncased \
+        --train_file {train_csv_no_text} \
+        --validation_file {val_csv} \
+        --text_column_names text \
+        --do_train \
+        --max_seq_length 32 \
+        --output_dir /tmp/test_output 2>&1
+    """
+    run_test(args, "文本列不存在，应抛出明确错误", expect_success=False)
+
+    # 3. 测试标签列缺失报错
+    print("\n\n【测试3】标签列缺失报错")
+    train_csv_no_label = create_test_csv('train_no_label.csv', [
+        ['text', 'wrong_label'],
+        ['Good movie', 'positive'],
+    ])
+
+    args = f"""--model_name_or_path distilbert-base-uncased \
+        --train_file {train_csv_no_label} \
+        --validation_file {val_csv} \
+        --label_column_name label \
+        --do_train \
+        --max_seq_length 32 \
+        --output_dir /tmp/test_output 2>&1
+    """
+    run_test(args, "标签列不存在，应抛出明确错误", expect_success=False)
+
+    # 4. 测试空文本处理 (应正常工作)
+    print("\n\n【测试4】空文本处理")
+    train_csv_empty = create_test_csv('train_empty_text.csv', [
+        ['text', 'label'],
+        ['', 'positive'],
+        [None, 'negative'],
+        ['normal text', 'positive'],
+    ])
+
+    args = f"""--model_name_or_path distilbert-base-uncased \
+        --train_file {train_csv_empty} \
+        --validation_file {val_csv} \
+        --do_train \
+        --per_device_train_batch_size 2 \
+        --num_train_epochs 1 \
+        --max_seq_length 32 \
+        --output_dir /tmp/test_output --overwrite_output_dir 2>&1
+    """
+    run_test(args, "空文本应正常处理，不崩溃", expect_success=True)
+
+    # 5. 验证标签一致性完整流程
+    print("\n\n【测试5】完整流程标签一致性")
+    print("正在训练基础模型...")
+
+    # 6. 运行单元测试
+    print("\n\n【测试6】运行单元测试验证修复逻辑")
+    result = subprocess.run('python test_label_mapping_unit.py', shell=True, cwd=test_dir, capture_output=True, text=True)
+    print(result.stdout)
+    if result.returncode == 0:
+        print("✅ 单元测试全部通过!")
+
+    print("\n" + "="*60)
+    print("边界情况测试完成!")
+    print("="*60)
+
+if __name__ == "__main__":
+    main()