Spaces:
Paused
Paused
| accelerate | |
| transformers | |
| sentencepiece | |
| pillow | |
| numpy | |
| torchvision | |
| huggingface_hub | |
| spaces | |
| opencv-python | |
| imageio | |
| imageio-ffmpeg | |
| einops | |
| timm | |
| av | |
| #flash-attn-3@https://huggingface.co/alexnasa/flash-attn-3/resolve/main/128/flash_attn_3-3.0.0b1-cp39-abi3-linux_x86_64.whl | |
| git+https://github.com/huggingface/diffusers.git@main | |
| Common | |
| # Configuration system. | |
| lpips>=0.1.4 # LPIPS model for VAE training | |
| # Dataloading | |
| bson>=0.5.10 # bson for Image dataloading | |
| tensorflow>=2.16.1 # Video dataloading | |
| opencv-python>=4.9.0.80 # OpenCV | |
| sentencepiece>=0.2.0 # For Text Encoder | |
| # Modeling | |
| rotary-embedding-torch>=0.5.3 # Rotary positional embedding | |
| transformers>=4.38.2 # Transformers | |
| torchvision>=0.19.0 # Torchvision | |
| tiktoken>=0.7.0 # Tiktoken for generation | |
| transformers_stream_generator>=0.0.5 #LLM generation support | |
| # Metrics | |
| torchmetrics>=1.3.2 # Core module for metric | |
| pycocotools>=2.0.7 # COCO-related | |
| torch-fidelity>=0.3.0 # FID-related | |
| # Experiment Tracking | |
| moviepy>=1.0.3 # WandB Logging Image & Video | |
| imageio>=2.34.0 # WandB Logging Image & Video | |
| tabulate>=0.9.0 # Logging Table | |
| deepdiff>=7.0.1 # Find difference of config | |
| # Testing | |
| parameterized>=0.9.0 # Define multiple tests through decorators. | |
| # Notebook | |
| mediapy>=1.2.0 # Notebook Visualization | |
| # DevOPs | |
| black >= 24 # Code formatting | |
| flake8 >= 7 # Code style | |
| isort >= 5 # Import sorting | |
| pre-commit>=3.7.0 # Pre-commit hooks | |
| expecttest>=0.2.1 # Pytorch dist tests | |
| hypothesis>=6.100.1 # Fix randomness | |
| av>=12.0.0 | |