ReplitLM

ReplitLM copied to clipboard

https://replit.com/@MATHANKUMARMK1/HTML-CSS-JS?s=app

Mathan-kumar-mk

This link is broken in the README: https://huggingface.co/spaces/replit/replit-code-v1-3b-demo

JohnAllen

Does ReplitLM support gradient checkpoints?

1

Does ReplitLM support gradient checkpoints?

sxthunder

I want to add FlashAttention2 to the replit-code-v1 for better performance and efficiency. Please, let me know If i am missing anything. @pirroh @madhavatreplit @amasad

kaushal07wick

Hi thanks for releasing this awesome codebase! I wonder if there is documentation on what results we should expect for running `bash scripts/multiple_eval.sh`.

ywen666

Can anyone have idea about my issue.

PuneethSaiCharanRaghavendra

Finetuning fails with `RuntimeError: Please install flash-attn==1.0.3.post0 and triton==2.0.0.dev20221202`

2

Following the instructions in the README, running through the docker container `mosaicml/llm-foundry:1.13.1_cu117-latest`, finetuning fails with `RuntimeError: Please install flash-attn==1.0.3.post0 and triton==2.0.0.dev20221202`. `pip install triton==2.0.0.dev20221202` fixes the problem. Ideally this hint...

matthiasgeihs

Speed up the inference

3

Hi, this model seems nice, but I do find that the inference speed is very slow (70ms/token on single A100), so I want to speed up it. It seems to...

Symbolk

How to train the ReplitLM model

1

When training, it is based on tramsformers' training course. It is started on the A100-80g machine, but the per gpu batch-size can be set to 2 at most, and there...

ran337287

Are there any vim plugins?

1

gunar