Hugging Face
Models
Datasets
Spaces
Posts
Docs
Enterprise
Pricing
Log In
Sign Up
THUDM
/
chatglm2-6b-int4
like
235
Follow
Knowledge Engineering Group (KEG) & Data Mining at Tsinghua University
1.33k
Transformers
PyTorch
Chinese
English
chatglm
glm
thudm
custom_code
Inference Endpoints
arxiv:
4 papers
Model card
Files
Files and versions
Community
24
Train
Deploy
Use this model
如果这个是用Bitsandsbyte的NF4量化的，能否直接在这个基础上用qlora继续训练？
#1
by
bash99
- opened
Jun 27, 2023
Discussion
bash99
Jun 27, 2023
还是建议在原始16bit模型上做训练然后再次量化？
另外这个量化是不是没有GPTQ推理加速的效果（对于llama模型非常明显）。
See translation
Edit
Preview
Upload images, audio, and videos by dragging in the text input, pasting, or
clicking here
.
Tap or paste here to upload images
Comment
·
Sign up
or
log in
to comment
