Hongyu Wang

Support for ChatGPT

1

I noticed that currently Jarvis only support text-DaVinci and gpt-4. Do you have any plan to support gpt-3.5-turbo in the near future? Cause it's more cheaper compared with current LLMs....

Training resources

3

Thanks for your wonderful work. I am very excited about your idea. May I ask the computation budget used to train the largest Imagebind model? How many GPU hour do...

Install apex without ninja

1

**Describe the Bug** I try to install apex following ***pip install -v --disable-pip-version-check --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./***. However, it reports ***subprocess.CalledProcessError: Command ‘[‘ninja‘, ‘-v‘]‘ returned non-zero exit status 1***. I...

Memory slowly increase as training progresses

9

Thanks for your wonderful work. I try to pre-train instrcutBLIP from scratch on 4x4 A100. However, the GPU memory is slowly increasing as the training progresses, which leads to OUT-OF-MEMORY...

Video request

4

Thanks for your wonderful work! I would like to download videos of VidChapters. Could you provide some cmd tools to quickly download these videos? And how much storage do you...

too many requests for download video from YouTube

2

Thanks for your work! I try to use video2dataset to download videos of VidChapter, a subset of YT-Temporal-1B, which contains 800k videos from YouTube. However, it reports too many request...

Thanks for your wonderful work. I try to use triton for mpt-1b-redpajama-200b. I use the default version of flash attention. However, it doesnot work. Do you meet the similar issues?...