Deema

icon indicating a website link https://deema-a.github.io/

icon company University of Illinois Urbana-Champaign icon location IL icon location Deema Nuhait => pronounced: No-Hate ;) \n This user runs on matcha, and ambitions

Results 11 comments of Deema

how to use peft at torch <=1.12.1 and CUDA <= 10.2 envrionment

I am not able to upgrade torch form 1.12.1 and wondering if any got this fixed?

Streaming audio data inference

@rosklyar @YuanxunLu may you please give some instructions/hints on how to make this streamable? Thanx!

Adding additional tokens to vocabulary

> When you add extra tokens, the embedding dimensions of some layers change. You have to add those layers in modules_to_save. Depending on your model, the modules_to_save can change. Please...

What is the meaning of implementing by C++?

I believe README needs more details on that

How to run inference of a (very) large model across mulitple GPUs ?

I am having the same question.

Does tensor_parallel support multi-node tensor parallel training?

same question.

[Assistance] Number of problems in the OS dataset

thank you @Longin-Yu apart form the `6-backup` the total is 182, including 26 dev. test 156 dev 26 but the stat here shows the test with 144 samples?

Documentation about setup_chat_format()

I have this issue: > size mismatch for base_model.model.model.embed_tokens.weight: copying a param with shape torch.Size([32002, 4096]) from checkpoint, the shape in current model is torch.Size([32000, 4096]). > size mismatch for...

Documentation about setup_chat_format()

> I've same issue too. Did you guys solve this problem? @deema-A @joaomsimoes Nope ='\

Support for PPO for PRM?

Hi, any updates on this?