Could someone please help?

[jeerychao]

Your question

Could someone please help? After I git cloned kijai's ComfyUI-FramePackWrapper in custom_nodes, I encountered an error when launching the workflow: "clip missing: ['text_projection.weight']". Additionally, when I upload images with descriptions, the generated video doesn't match my expectations.

My environment:

Ubuntu 22.04

Python 3.10.6

PyTorch version: 2.6.0+cu124

Model directory setup (referenced from GitHub):

/apps/ComfyUI/models/clip_vision: Contains sigclip_vision_patch14_384.safetensors

/apps/ComfyUI/models/text_encoders: Contains llava_llama3_fp16.safetensors, llava_llama3_fp8_scaled.safetensors, and clip_l.safetensors

/apps/ComfyUI/models/vae: Contains hunyuan_video_vae_bf16.safetensors

/apps/ComfyUI/models/diffusers/lllyasviel/FramePackI2V_HY/: Contains 3 diffusion_pytorch_model-0000x-of-00003.safetensors files, diffusion_pytorch_model.safetensors.index.json, and config.json.

Workflow file: framepack_hv_example.json.

framepack_hv_example.json

Logs

Other

No response

[jeerychao]

help me,please

[Buumcode]

You can consult with ChatGPT -- he is strong in this things. I've use this one (free, no registration) - https://lmarena.ai/ When loading, press OK for agreeing terms of service, in bottom (near clip) -- ask what you want (on your native language). ChatGPT did answer!

[github-actions[bot]]

This issue is being marked stale because it has not had any activity for 30 days. Reply below within 7 days if your issue still isn't solved, and it will be left open. Otherwise, the issue will be closed automatically.