Poorly Written requirements.txt

[samarth777]

The provided requirements.txt for the StableVideo model lacks crucial dependencies and contains incorrect version specifications, leading to installation failures. Many essential packages required for the model's proper functioning are missing, and users are forced to manually search and install them, along with downgrading certain dependencies to make the model run successfully.

[computersrmyfriends]

Could you please share the updated requirements.txt?

[rese1f]

try this version: dependencies:

• python=3.8.5
• pip=20.3
• cudatoolkit=11.3
• pytorch=1.12.1
• torchvision=0.13.1
• numpy=1.23.1
• pip:
  • gradio==3.16.2
  • albumentations==1.3.0
  • opencv-contrib-python==4.3.0.36
  • imageio==2.9.0
  • imageio-ffmpeg==0.4.2
  • pytorch-lightning==1.5.0
  • omegaconf==2.1.1
  • test-tube>=0.7.5
  • streamlit==1.12.1
  • einops==0.3.0
  • transformers==4.19.2
  • webdataset==0.2.5
  • kornia==0.6
  • open_clip_torch==2.0.2
  • invisible-watermark>=0.1.5
  • streamlit-drawable-canvas==0.8.0
  • torchmetrics==0.6.0
  • timm==0.6.12
  • addict==2.4.0
  • yapf==0.32.0
  • prettytable==3.6.0
  • safetensors==0.2.7
  • basicsr==1.4.2

[zrthxn]

It looks like the author's whole environment was just dumped into a requirements.txt with pip freeze, packages that are not being used at all were there.

After spending a lot of time fixing issues with the requirements, this minimal set of packages works for me to run this.

torch
numpy
einops
pytorch_lightning
timm
omegaconf
open_clip_torch[training]
opencv-python
scipy
imageio[ffmpeg]

I've removed the requirement of the package basicsr which was used only in one place to download a file! I have also had to make tons of changes to the code to fix simple errors, if anyone is interested you can see my fork

I've tested this on CPU (Apple M2) and Colab V100 and A100 GPUs.