Jiarui Fang（方佳瑞）
Jiarui Fang（方佳瑞）
@MikeChenfu I have updated the OPT example code. And provide detailed benchmark results for the GPT example. I suppose OPT and GPT have similar performances. How much CPU memory did...
Thanks @eric8607242 , very impressive. Could you please solve the conflict? @1SAA please help review this PR.
@haofanwang sorry, it is a bug could you please install CAI using the following cmd. ``` CUDA_EXT=1 pip install . ```
@haofanwang It has worked already. You can ignore those warnings. Anyway, we will fix them later.
Hey, ColossalAI-Examples have been deprecated. You can refer to the examples in ColossalAI/example. Although there is no resnet example right now, you can build it on your own. Looking forwards...
Did you install it from our official website. If install from source, please check if your CUDA tookit version matches your torch and colossalai version.
The biggest question in my mind is: How to initialize optimizer states when finetuning a pre-trained model?
> Hi, I found this [picture](https://raw.githubusercontent.com/hpcaitech/public_assets/main/applications/chatgpt/experience.jpg) a little confusing. I think if there is some description, it would be much better. Yes, at least tell the readers: the green part...
Thanks for your insights. Can you help with this project? From my point of view, I have not seen real requirements of pipeline parallelism for U-net. So it is not...
@haofanwang correct! `get_static_torch_model ` will return a torch model, you can do whatever you want on it. @1SAA @FrankLeeeee we should add a chapter in our document on model saving...