lllyasviel / lllyasviel/FramePack
Abnormal memory usage
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 17.3k
- Forks
- 1.7k
- PR merge metrics
- No merged PRs in 30d
Description
People here say they ran it on 6GB VRAM without any problems...
My spec: 4070 SUPER 12 Gb + 32RAM Windows 11
Problems start at begining, i need to increase my virtual memory to 32Gb and close everything i can just to launch it and my RAM still almost empty... Otherwise python just crash on shards loading.
In second i cant generate anything with any setting. When i start, my RAM and virtual RAM fills at 100% and my VRAM only ~15%. After my pc struggling for a bit i see this:
```
Unloaded DynamicSwap_LlamaModel as complete.
Unloaded CLIPTextModel as complete.
Unloaded SiglipVisionModel as complete.
Unloaded AutoencoderKLHunyuanVideo as complete.
Unloaded DynamicSwap_HunyuanVideoTransformer3DModelPacked as complete.
Loaded CLIPTextModel to cuda:0 as complete.
Unloaded CLIPTextModel as complete.
Loaded AutoencoderKLHunyuanVideo to cuda:0 as complete.
Unloaded AutoencoderKLHunyuanVideo as complete.
Loaded SiglipVisionModel to cuda:0 as complete.
latent_padding_size = 27, is_last_section = False
Unloaded SiglipVisionModel as complete.
Moving DynamicSwap_HunyuanVideoTransformer3DModelPacked to cuda:0 with preserved memory: 8 GB
0%| | 0/25 [00:02
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reproducing the reported generation failure through webui/demo_gradio.py, then trace the call into webui/diffusers_helper/k_diffusion/k_diffusion_hunyuan.py and webui/diffusers_helper/models/hunyuan_video_packed.py. Compare the reported CUDA out-of-memory point and memory usage with a working configuration; done means the cause is identified and generation no longer fails under the stated setup.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning, performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100