Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

加入 ChatRWKV 支持,请问开发者有没有联系方式 #47

Open
BlinkDL opened this issue Apr 2, 2023 · 5 comments
Open

加入 ChatRWKV 支持,请问开发者有没有联系方式 #47

BlinkDL opened this issue Apr 2, 2023 · 5 comments

Comments

@BlinkDL
Copy link

BlinkDL commented Apr 2, 2023

Hi 大家好,我是 RWKV 的作者,目前有中英文 Chat 模型和小说模型,7B 和 14B:

https://zhuanlan.zhihu.com/p/618011122

RWKV 现在有 pip package 可以直接调用推理,支持 INT8 量化,支持 streaming 模式(可以用很小显存运行),支持拆分到多张显卡:

https://pypi.org/project/rwkv/

大家可以合作加入 RWKV 支持吗?如有兴趣可以加 RWKV 的 QQ 群,谢谢。请问 ChatGLM 有没有群,我也加。

@roj234
Copy link

roj234 commented Apr 2, 2023

我不是作者。

在huggingface的仓库里写config.json,然后就可以加参数--model BlinkDL/rwkv-4-raven,而不需要这边做什么改动
因为用的是AutoModel.from_pretrained
举例
{
"_name_or_path": "THUDM/chatglm-6b",
"architectures": [
"ChatGLMModel"
],
"auto_map": {
"AutoConfig": "configuration_chatglm.ChatGLMConfig",
"AutoModel": "modeling_chatglm.ChatGLMForConditionalGeneration",
"AutoModelForSeq2SeqLM": "modeling_chatglm.ChatGLMForConditionalGeneration"
},
"bos_token_id": 150004,
"eos_token_id": 150005,
"hidden_size": 4096,
"inner_hidden_size": 16384,
"layernorm_epsilon": 1e-05,
"max_sequence_length": 2048,
"model_type": "chatglm",
"num_attention_heads": 32,
"num_layers": 28,
"position_encoding_2d": true,
"torch_dtype": "float16",
"transformers_version": "4.23.1",
"use_cache": false,
"vocab_size": 150528
}
和ChatGLM项目的其余几个py

@BlinkDL
Copy link
Author

BlinkDL commented Apr 2, 2023

RWKV不是GPT模型,目前也还没有加入Huggingface transformers(社区在做,但进度缓慢)

所以目前最方便是调用 RWKV pip package

@wfjsw
Copy link

wfjsw commented Apr 2, 2023

@l15y
Copy link

l15y commented Apr 3, 2023

https://github.com/l15y/wenda

@roj234
Copy link

roj234 commented Apr 7, 2023

https://github.com/roj234/ChatGLM-webui/tree/rwkv

话说rwkv的PIPELINE的sample_logits不知道为什么很卡
还有猫娘真的很哲学,而且出现概率并不是很低

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants