| numpy>=1.26.4 | |
| # to quote a-r-r-o-w/finetrainers: | |
| # It is recommended to use Pytorch 2.5.1 or above for training. Previous versions can lead to completely black videos, OOM errors, or other issues and are not tested. | |
| # on some system (Python 3.13+) those do not work: | |
| torch==2.5.1 | |
| torchvision==0.20.1 | |
| torchao>=0.7.0 | |
| # datasets 3.4.0 replaces decord by torchvision | |
| # let's free it for now | |
| datasets==3.3.2 | |
| huggingface_hub | |
| hf_transfer>=0.1.8 | |
| diffusers @ git+https://github.com/huggingface/diffusers.git@main | |
| transformers>=4.45.2 | |
| accelerate | |
| bitsandbytes | |
| peft>=0.12.0 | |
| # For GPU monitoring of NVIDIA chipsets | |
| you probably won't be able to install that on macOS | |
| # pynvml | |
| # eva-decord is missing get_batch it seems | |
| eva-decord==0.6.1 | |
| # decord | |
| finetrainers @ git+https://github.com/a-r-r-o-w/finetrainers.git@4c6844fabf496fb622b6333f57e8ebecee6f2780 | |
| wandb | |
| pandas | |
| sentencepiece>=0.2.0 | |
| imageio | |
| imageio-ffmpeg | |
| torchdata==0.11.0 | |
| # for youtube video download | |
| pytube | |
| pytubefix | |
| # for scene splitting | |
| scenedetect[opencv] | |
| # for llava video / captionning | |
| pillow | |
| pillow-avif-plugin | |
| polars | |
| einops | |
| open_clip_torch | |
| av==14.1.0 | |
| git+https://github.com/LLaVA-VL/LLaVA-NeXT.git | |
| # for our frontend | |
| gradio==5.20.1 | |
| gradio_toggle | |
| # used for the monitor | |
| matplotlib |