tensorflow版本的模型 ：
链接：https://pan.baidu.com/s/10tjVfypoQy6G_mkZK6cqOQ?pwd=yljr
提取码：yljr
模型文件名 模型简称 语料 版本 地址 github地址 加载方式 huggingface加载
chinese_roberta_wwm_large_ext_L-24_H-1024_A-16 RoBERTa-wwm-ext-large, Chinese 中文维基百科，其他百科、新闻、问答等数据，总词数达5.4B tensorflow 百度网盘 请输入提取码 GitHub - ymcui/Chinese-BERT-wwm: Pre-Training with Whole Word Masking for Chinese BERT（中文BERT-wwm系列模型）
hflchinese-roberta-wwm-ext-large pytorch https://huggingface.co/hfl/chinese-roberta-wwm-ext-large hfl/chinese-roberta-wwm-ext-large
chinese_roberta_wwm_ext_L-12_H-768_A-12 RoBERTa-wwm-ext, Chinese tensorflow 百度网盘 请输入提取码
hflchinese-roberta-wwm-ext pytorch https://huggingface.co/hfl/chinese-roberta-wwm-ext hfl/chinese-roberta-wwm-ext
chinese_bert_wwm_ext_L-12_H-768_A-12 BERT-wwm-ext, Chinese tensorflow 百度网盘 请输入提取码
hflchinese-bert-wwm-ext pytorch hfl/chinese-bert-wwm-ext · Hugging Face hfl/chinese-bert-wwm-ext
chinese_bert_wwm_L-12_H-768_A-12 BERT-wwm, Chinese tensorflow 百度网盘 请输入提取码
hfl/chinese-bert-wwm-ext pytorch hfl/chinese-bert-wwm
GitHub - brightmart/roberta_zh: RoBERTa中文预训练模型: RoBERTa for Chinese
roberta_zh_l12 RoBERTa_zh_L12 30G原始文本，近3亿个句子，100亿个中文字(token)，产生了2.5亿个训练数据(instance)；覆盖新闻、社区问答、多个百科数据等； tensorflow roberta_zh_l12.zip_免费高速下载|百度网盘-分享无限制 Bert 直接加载
roeberta_zh_L-24_H-1024_A-16 RoBERTa-zh-Large tensorflow roeberta_zh_L-24_H-1024_A-16.zip_免费高速下载|百度网盘-分享无限制 Bert 直接加载
RoBERTa_zh_L12_PyTorch RoBERTa_zh_L12 pytorch RoBERTa_zh_L12_PyTorch.zip_免费高速下载|百度网盘-分享无限制 Bert的PyTorch版直接加载
chinese_L-12_H-768_A-12 bert_base, Chinese 中文维基百科 https://storage.googleapis.com/bert_models/2018_11_03/chinese_L-12_H-768_A-12.zip GitHub - google-research/bert: TensorFlow code and pre-trained models for BERT
uncased_L-2_H-128_A-2 bert_tiny, 24个bert_uncased模型 https://storage.googleapis.com/bert_models/2020_02_20/all_bert_models.zip
uncased_L-4_H-256_A-4 bert_mini, 24个bert_uncased 模型
uncased_L-4_H-512_A-8 bert_small,24个bert_uncased
uncased_L-8_H-512_A-8 bert_medium, 24个bert_uncased
uncased_L-12_H-768_A-12 bert_base, 24个bert_uncased
chinese_xlnet_mid_L-24_H-768_A-12 XLNet-mid, Chinese 中文维基百科，其他百科、新闻、问答等数据，总词数达5.4B tensorflow 百度网盘 请输入提取码 GitHub - ymcui/Chinese-XLNet: Pre-Trained Chinese XLNet（中文XLNet预训练模型） hfl/chinese-xlnet-mid
hflchinese-xlnet-mid pytorch https://huggingface.co/hfl/chinese-xlnet-mid
chinese_xlnet_base_L-12_H-768_A-12 XLNet-base, Chinese tensorflow 百度网盘 请输入提取码 hfl/chinese-xlnet-base
hflchinese-xlnet-base pytorch https://huggingface.co/hfl/chinese-xlnet-base
albert_tiny_zh albert_tiny_zh https://storage.googleapis.com/albert_zh/albert_tiny.zip GitHub - brightmart/albert_zh: A LITE BERT FOR SELF-SUPERVISED LEARNING OF LANGUAGE REPRESENTATIONS, 海量中文预训练ALBERT模型
albert_tiny_489k albert_tiny_zh(训练更久，累积学习20亿个样本) https://storage.googleapis.com/albert_zh/albert_tiny_489k.zip
albert_tiny_zh_google albert_tiny_google_zh(累积学习10亿个样本,google版本) https://storage.googleapis.com/albert_zh/albert_tiny_zh_google.zip
albert_small_zh_google albert_small_google_zh(累积学习10亿个样本,google版本) https://storage.googleapis.com/albert_zh/albert_small_zh_google.zip
albert_large_zh albert_large_zh https://storage.googleapis.com/albert_zh/albert_large_zh.zip
albert_base_zh albert_base_zh(小模型体验版) https://storage.googleapis.com/albert_zh/albert_base_zh.zip
albert_base_zh_additional_36k_steps albert_base_zh(额外训练了1.5亿个实例即 36k steps * batch_size 4096) https://storage.googleapis.com/albert_zh/albert_base_zh_additional_36k_steps.zip
albert_xlarge_zh_177k albert_xlarge_zh_177k https://storage.googleapis.com/albert_zh/albert_xlarge_zh_177k.zip
albert_xlarge_zh_183k albert_xlarge_zh_183k（优先尝试） https://storage.googleapis.com/albert_zh/albert_xlarge_zh_183k.zip
voidfulalbert_chinese_tiny pytorch https://huggingface.co/voidful/albert_chinese_tiny voidful/albert_chinese_tiny
voidfulalbert_chinese_small pytorch https://huggingface.co/voidful/albert_chinese_small voidful/albert_chinese_small
voidfulalbert_chinese_base pytorch voidful/albert_chinese_base · Hugging Face voidful/albert_chinese_base
voidfulalbert_chinese_large pytorch https://huggingface.co/voidful/albert_chinese_large voidful/albert_chinese_large
voidfulalbert_chinese_xlarge pytorch voidful/albert_chinese_xlarge · Hugging Face voidful/albert_chinese_xlarge
voidfulalbert_chinese_xxlarge pytorch https://huggingface.co/voidful/albert_chinese_xxlarge voidful/albert_chinese_xxlarge
electra_180g_large ELECTRA-180g-large, Chinese tensorflow 百度网盘 请输入提取码 GitHub - ymcui/Chinese-ELECTRA: Pre-trained Chinese ELECTRA（中文ELECTRA预训练模型） hfl/chinese-electra-180g-large-discriminator
hflchinese-electra-180g-large-discriminator pytorch https://huggingface.co/hfl/chinese-electra-180g-large-discriminator
electra_180g_base ELECTRA-180g-base, Chinese tensorflow 百度网盘 请输入提取码 hfl/chinese-electra-180g-base-discriminator
hflchinese-electra-180g-base-discriminator pytorch https://huggingface.co/hfl/chinese-electra-180g-base-discriminator
electra_180g_small_ex ELECTRA-180g-small-ex, Chinese tensorflow 百度网盘 请输入提取码 hfl/chinese-electra-180g-small-ex-discriminator
hflchinese-electra-180g-small-ex-discriminator pytorch https://huggingface.co/hfl/chinese-electra-180g-small-ex-discriminator
electra_180g_small ELECTRA-small, Chinese tensorflow 百度网盘 请输入提取码 hfl/chinese-electra-180g-small-discriminator
hflchinese-electra-180g-small-discriminator pytorch https://huggingface.co/hfl/chinese-electra-180g-small-discriminator
chinese_macbert_large MacBERT-large, Chinese tensorflow 百度网盘 请输入提取码 GitHub - ymcui/MacBERT: Revisiting Pre-trained Models for Chinese Natural Language Processing (MacBERT) hfl/chinese-macbert-large
hflchinese-macbert-large pytorch https://huggingface.co/hfl/chinese-macbert-large
chinese_macbert_base MacBERT-base, Chinese tensorflow 百度网盘 请输入提取码 hfl/chinese-macbert-base
hflchinese-macbert-base pytorch https://cdn-lfs.huggingface.co/hfl/chinese-macbert-base
Erlangshen-SimCSE-110M-Chinese pytorch
unsup-simcse-bert-base-uncased pytorch
sup-simcse-roberta-large pytorch
sup-simcse-bert-base-uncased pytorch
albert-base-chinese-cluecorpussmall pytorch
# bert # 语言模型 # xlnet # 自然语言处理
Logo
2048 AI社区

有“AI”的1024 = 2048，欢迎大家加入2048 AI社区

更多推荐

cover

Python Django 疫情数据可视化分析系统平台完整项目源码+SQL脚本+接口文档【Django Web毕设】

avatar 2048 AI社区

产业联盟应如何借助人工智能来盘活产学研合作的产业竞争力？

AI赋能产学研合作新生态：破解科技成果转化瓶颈 摘要：在科技竞争加剧的背景下，产业联盟作为产学研合作载体面临信息不对称、供需匹配低效等挑战。AI技术为破解这些难题提供了新思路：通过大数据分析实现精准匹配，利用自然语言处理挖掘技术价值，构建智能评估模型判断专利潜力。实践证明，AI驱动的专利价值评估、企业需求挖掘等场景显著提升了合作效率。构建基于AI的开放生态需要政府、企业、高校和产业联盟协同创新，实

avatar 2048 AI社区

用AI生成的html页面设计放到到Asure上实现再改造的方法

【摘要】AI生成的HTML原型导入Axure需通过Figma转换，具体步骤：1）在Figma安装"htmltodesign"插件导入HTML文件；2）安装Axure插件复制原型内容；3）Axure中粘贴并编辑原型。注意事项包括：确保HTML文件完整、处理样式错乱问题、检查外部文件路径等。该方法适用于需在Axure中进一步编辑的场景，若仅使用Figma设计可直接跳过Axure步骤

avatar 2048 AI社区