常用的中文预训练模型、预训练词向量下载地址收藏

中文预训练模型下载

原版:https://huggingface.co/models

Google原版bert:https://github.com/google-research/bert

  • 中文XLNET预训练模型:https://github.com/ymcui/Chinese-XLNet (只有TensorFlow需转化) or https://huggingface.co/hfl/chinese-xlnet-base/tree/main (pytorch版本需下载)可参照:https://www.bilibili.com/read/cv8231417/

  • 中文BERT预训练模型:https://storage.googleapis.com/bert_models/2018_11_03/chinese_L-12_H-768_A-12.zip

  • 中文MacBERT预训练模型:https://github.com/ymcui/MacBERT

  • 中文ELECTRA预训练模型:https://github.com/ymcui/Chinese-ELECTRA

  • 中文BERT-wwm预训练模型:https://github.com/ymcui/Chinese-BERT-wwm

  • 中文RoBERTa预训练模型:https://github.com/brightmart/roberta_zh

  • 知识蒸馏工具TextBrewer:https://github.com/airaria/TextBrewer

中文预训练词向量下载

常用英文单词的词向量:

  • glove.6B.zip:里面包含了50d,100d,200d,300d常用英文单词的词向量

    链接:https://pan.baidu.com/s/1m5zKaJGFwV1VNTsHgGqwRw
    提取码:5knd

常用中文词向量:https://github.com/Embedding/Chinese-Word-Vectors

  • merge_sgns_bigram_char300.txt.bz2:Mixed-large 综合 Baidu Netdisk / Google Drive, Word + Character + Ngram

  • sgns.baidubaike.bigram-char.bz2:Baidu Encyclopedia 百度百科, Word + Character + Ngram

    https://pan.baidu.com/s/1IiIbQGJ_AooTj5s8aZYcvA

    PWD: 5555

欢迎各位关注我的个人公众号:HsuDan,我将分享更多自己的学习心得、避坑总结、面试经验、AI最新技术资讯。

你可能感兴趣的:(NLP,深度学习,AI,自然语言处理,深度学习,tensorflow)