解决Colab中ModuleNotFoundError: No module named 'transformers.models.mmbt'报错

问题背景 #

在Google Colab运行基于classla/xlm-roberta-base-multilingual-text-genre-classifier的文本分类代码时，突然出现上述报错，此前代码可正常运行。当前环境：

• transformers 4.30.2
• simpletransformers 0.63.11
• Python 3

原因 #

simpletransformers 0.63.11版本的多模态分类模块依赖transformers.models.mmbt，但transformers从4.29.0版本开始，已将MMBT（多模态双向Transformer）从核心库中移除，移至单独的扩展组件。

解决方案 #

方案1：安装transformers的MMBT扩展组件 #

在Colab中执行以下命令，重新安装带MMBT支持的指定版本transformers：

!pip uninstall -y transformers
!pip install transformers[mmbt]==4.30.2

方案2：降级transformers到仍包含MMBT的版本 #

如果方案1无效，可降级transformers到4.28.1版本（该版本仍将MMBT保留在核心库中）：

!pip uninstall -y transformers
!pip install transformers==4.28.1

方案3：绕过simpletransformers，直接使用transformers核心库 #

因为实际只用到XLM-Roberta的文本分类功能，完全可以不用simpletransformers，直接用transformers原生接口实现，避免依赖问题。替换代码如下：

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

# 加载模型和分词器
tokenizer = AutoTokenizer.from_pretrained("classla/xlm-roberta-base-multilingual-text-genre-classifier")
model = AutoModelForSequenceClassification.from_pretrained("classla/xlm-roberta-base-multilingual-text-genre-classifier")
model.to("cuda")  # 使用GPU

# 待预测文本
texts = [
    "How to create a good text classification model? First step is to prepare good data. Make sure not to skip the exploratory data analysis. Pre-process the text if necessary for the task. The next step is to perform hyperparameter search to find the optimum hyperparameters. After fine-tuning the model, you should look into the predictions and analyze the model's performance. You might want to perform the post-processing of data as well and keep only reliable predictions.",
    "On our site, you can find a great genre identification model which you can use for thousands of different tasks. With our model, you can fastly and reliably obtain high-quality genre predictions and explore which genres exist in your corpora. Available for free!"
]

# 预处理文本
inputs = tokenizer(texts, padding=True, truncation=True, max_length=512, return_tensors="pt").to("cuda")

# 预测
with torch.no_grad():
    outputs = model(**inputs)
predictions = torch.argmax(outputs.logits, dim=-1).cpu().numpy()

# 获取标签
labels = [model.config.id2label[i] for i in predictions]
print(predictions)  # 输出: [3 8]
print(labels)       # 输出: ['Instruction', 'Promotion']

内容的提问来源于stack exchange，提问作者simKO