YAML Metadata Warning: The pipeline tag "text2text-generation" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other

只需要700MB就可以运行的模型

如何判断电脑是否可以运行

打开B主页,点进去一个视频;如果这个时候你的电脑没有爆炸,就可以运行。

介绍

使用了MiniMind2的模型参数:

快速开始

安装依赖:

pip install torch, transformer

运行模型:

python model_congnilite.py

常见问题介绍

在流式输出中,每输出一个token_id,就将它解码为字符并输出,会造成中文乱码现象,但是将token_id放到一个列表中一起解码就不会出现乱码

专业描述:token边界不对齐导致的解码错误

  • tokenizer采用的是子词(subword)分词(如BPE、SentencePiece等),一个汉字或词语可能被拆成多个token。
  • 单独解码一个token_id时,tokenizer.decode()会把这个token当作一个完整的单元去还原为字符,但实际上它可能只是一个汉字的“片段”或“字节”,导致输出乱码或不可见字符。
  • 只有把一组token_id(即一个完整的token序列)一起decode,tokenizer才能正确地拼接还原出原始的中文字符。

原本的代码:

new_token_str = tokenizer.decode(next_token_id.item(), skip_special_tokens=False)
print(new_token_str, end='', flush=True)

更改后:

prev_decoded = tokenizer.decode(token_list[:-1], skip_special_tokens=False)
curr_decoded = tokenizer.decode(token_list, skip_special_tokens=False)
print(curr_decoded[len(prev_decoded):], end='', flush=True)
Downloads last month
-
Safetensors
Model size
109M params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support