Exception: Failed to init class <class 'tools.modules.clip_embedder.FrozenOpenCLIPTtxtVisualEmbedder'>, with Pretrained weights (models/open_clip_pytorch_model.bin) not found for model ViT-H-14.

[A-Salty-Fish]

RuntimeError: Pretrained weights (models/open_clip_pytorch_model.bin) not found for model ViT-H-14.

I have runned the following code successfully: from modelscope.hub.snapshot_download import snapshot_download model_dir = snapshot_download('damo/I2VGen-XL', cache_dir='models/')

[Steven-SWZhang]

Have you placed the downloaded models in the "models" directory?

[sayakpaul]

It still happens. My models directory has:

doc                          README.md  stable_diffusion_image_key_temporal_attention_x1.json
i2vgen_xl_00854500.pth       README.MD  v2-1_512-ema-pruned.ckpt
open_clip_pytorch_model.bin  source

Now, when I run:

python inference.py --cfg configs/i2vgen_xl_infer.yaml

It results in the error.

[himali94]

Same error Exception: Failed to init class <class 'tools.modules.clip_embedder.FrozenOpenCLIPTtxtVisualEmbedder'>, with Model config for VIT-H-14 not found.

[LJ-lijian]

I got the same error, did someone solve this problem?

[wuguowuge]

I got the same error +1, did someone solve this problem?

[zhenglinpan]

I cloned original ViT-H-14 from huggingface which provides both model & config file, merged with model folder, looked fine tho.

Hope repository owner make this clear.

[h-pal]

@zhenglinpan I cloned the same VIT-H-14 from huggingface but im getting Exception: Failed to invoke function <function train_t2v_entrance at 0x151a326d6af0>, with Failed to init class <class 'tools.modules.clip_embedder.FrozenOpenCLIPTextVisualEmbedder'>, with invalid load key, 'v'. any help is appreciated