Error with LLM text encoder (has autodownload):

[tcla75]

Downloading clip model to: C:\SD\ComfyUI2\New\ComfyUI_windows_portable\ComfyUI\models\clip\clip-vit-large-patch14
Fetching 10 files: 40%|█████████████████████████▌ | 4/10 [01:02<01:33, 15.61s/it]
!!! Exception during processing !!! Consistency check failed: file should be of size 1710540580 but has size 568551292 (model.safetensors).

[kijai]

Huggingface downloader sometimes has these issues, you can also manually get the files from here:

https://huggingface.co/Kijai/llava-llama-3-8b-text-encoder-tokenizer/tree/main

to:
ComfyUI/models/LLM/llava-llama-3-8b-text-encoder-tokenizer

[chgxtony]

Error(s) in loading state_dict for CLIPTextModel:
size mismatch for text_model.embeddings.token_embedding.weight: copying a param with shape torch.Size([49408, 768]) from checkpoint, the shape in current model is torch.Size([49408, 512]).

[doogyhatts]

I got this error too for DownloadAndLoadHyVideoTextEncoder

Error(s) in loading state_dict for CLIPTextModel:
size mismatch for text_model.embeddings.token_embedding.weight: copying a param with shape torch.Size([49408, 768]) from checkpoint, the shape in current model is torch.Size([49408, 512])

I did a manual download of the file using wget instead of relying on the auto-download.

[zhao-kun]

I got this error too for DownloadAndLoadHyVideoTextEncoder

Error(s) in loading state_dict for CLIPTextModel: size mismatch for text_model.embeddings.token_embedding.weight: copying a param with shape torch.Size([49408, 768]) from checkpoint, the shape in current model is torch.Size([49408, 512])

I did a manual download of the file using wget instead of relying on the auto-download.

I encountered the same issue. I also manually downloaded model.safetensors from the https://huggingface.co/openai/clip-vit-large-patch14 and placed it in the ComfyUI/models/clip/clip-vit-large-patch14 directory.

[doogyhatts]

The solution is found here.
#13