mirror of
https://github.com/Wan-Video/Wan2.1.git
synced 2025-11-04 14:16:57 +00:00
62 lines
855 B
Plaintext
62 lines
855 B
Plaintext
# Core AI stack
|
|
diffusers==0.34.0
|
|
transformers==4.53.1
|
|
tokenizers>=0.20.3
|
|
accelerate>=1.1.1
|
|
tqdm
|
|
imageio
|
|
imageio-ffmpeg
|
|
einops
|
|
sentencepiece
|
|
open_clip_torch>=2.29.0
|
|
|
|
# Video & media
|
|
moviepy==1.0.3
|
|
av
|
|
ffmpeg-python
|
|
pygame>=2.1.0
|
|
sounddevice>=0.4.0
|
|
soundfile
|
|
mutagen
|
|
pyloudnorm
|
|
librosa==0.11.0
|
|
|
|
# UI & interaction
|
|
gradio==5.23.0
|
|
dashscope
|
|
loguru
|
|
|
|
# Vision & segmentation
|
|
opencv-python>=4.9.0.80
|
|
segment-anything
|
|
rembg[gpu]==2.0.65
|
|
onnxruntime-gpu
|
|
decord
|
|
timm
|
|
|
|
# Config & orchestration
|
|
omegaconf
|
|
hydra-core
|
|
easydict
|
|
pydantic==2.10.6
|
|
|
|
# Math & modeling
|
|
torchdiffeq>=0.2.5
|
|
tensordict>=0.6.1
|
|
mmgp==3.5.10
|
|
peft==0.17.0
|
|
matplotlib
|
|
|
|
# Utilities
|
|
ftfy
|
|
piexif
|
|
pynvml
|
|
misaki
|
|
|
|
# Optional / commented out
|
|
# transformers==4.46.3 # for llamallava pre-patch
|
|
# rembg==2.0.65 # non-GPU fallback
|
|
# huggingface_hub[hf_xet] # slows down everything
|
|
# num2words
|
|
# spacy
|