binmakeswell comments

Results 359 comments of


                                            binmakeswell

[DOC]: 说明文档太少了啊，对新手不友好

Hi @infoBrainSys What part of the documentation are you looking for? Could you please provide details? We are in a high frequency of updates. Thanks. General Doc https://colossalai.org/ General example...

[FEATURE]: serving multiple models

Hi @dlzou Yes, we are considering it, https://github.com/orgs/hpcaitech/projects/17/views/1

[BUG]: The training code of reward model may be wrong

Hi @Luoyang144 @LuciusMos @MyHerbTea @YiAthena After verification, this is not a bug caused by the code, but an inappropriate sh command. We have fixed it. Thanks. https://github.com/hpcaitech/ColossalAI/blob/main/applications/Chat/examples/train_rm.sh

[BUG]: Is it normal to have loss nan after the Stage 1 - Supervised Finetuning?

Thanks for the feedback, we are fixing LoRA bug.

[BUG]: training GPT2-S using a single card on colab, AssertionError: You should use `zero_ddp_wrapper` first

Hi @LivinLuo1993 We did not conduct testing on Colab, and a Linux server is the preferred choice. Sorry about it.

Compativility for amd

Received, will consider in follow-up. Thanks for the feedback.

[BUG]: Error building extension 'cpu_adam'

Hi @Tian14267 The possible reason is GCC version is too low, or env is not compatible. Please check https://github.com/hpcaitech/ColossalAI#installation Thanks.

Can not train llama-7b due to OOM on 40GA100

Hi guys, we will double-check it this week and give a detailed example for a limited resource. Thanks.

Tensor并行的作用范围？

Hi @bobo0810 Thanks for the suggestion, we will add it in the near future when we update the documentation.

[BUG]: how can I run ChatGPT via inference.py which is at the example file?

Hi @hongkyunn You can refer to https://github.com/hpcaitech/ColossalAI/tree/main/applications/ChatGPT/examples#inference-exampleafter-stage3