lllyasviel / lllyasviel/FramePack
First round OK, second round 100% CPU and 100% RAM usage
- Dominant language
- Python
- Stars
- 17.3k
- Forks
- 1.7k
- PR merge metrics
- No merged PRs in 30d
Description
I installed in a aws instance : nvida L40 , 46G vRAM and 30G CPU RAM.
first round is OK to generate video in gradio UI, then click button to generate the second video for same input in gradio UI, server will get 100% CPU and 100% RAM usage, but GPU is OK , using 32G / 46G vRAM .
I added my memory clean up login into the [demo_gradio.py](https://github.com/lllyasviel/FramePack/blob/main/demo_gradio.py), did not get improved.
There is memory leaking somewhere in the model.
Contributor guide
No contributing guide indexed for this repository
Research direction
Start with demo_gradio.py and reproduce a second video generation with the same input in the Gradio UI on the reported AWS instance. Track CPU and system RAM alongside GPU memory, and inspect the existing cleanup logic mentioned in the issue. Done means repeated generations no longer drive CPU and RAM to 100% while GPU use remains stable.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 32/100