pytorch-image-models
pytorch-image-models copied to clipboard
Cannot create model eva02_enormous_patch14_clip_224.pretrain
I encounter following error with timm 0.9.1
Traceback (most recent call last):
File "download_models.py", line 63, in <module>
model = timm.create_model('eva02_enormous_patch14_clip_224.pretrain', pretrained=True)
File "/home/luoxin/.local/lib/python3.8/site-packages/timm/models/_factory.py", line 114, in create_model
model = create_fn(
File "/home/luoxin/.local/lib/python3.8/site-packages/timm/models/eva.py", line 1084, in eva02_enormous_patch14_clip_224
model = _create_eva('eva02_enormous_patch14_clip_224', pretrained=pretrained, **dict(model_args, **kwargs))
File "/home/luoxin/.local/lib/python3.8/site-packages/timm/models/eva.py", line 645, in _create_eva
model = build_model_with_cfg(
File "/home/luoxin/.local/lib/python3.8/site-packages/timm/models/_builder.py", line 393, in build_model_with_cfg
load_pretrained(
File "/home/luoxin/.local/lib/python3.8/site-packages/timm/models/_builder.py", line 189, in load_pretrained
raise RuntimeError(f"No pretrained weights exist for {model_name}. Use `pretrained=False` for random init.")
RuntimeError: No pretrained weights exist for eva02_enormous_patch14_clip_224. Use `pretrained=False` for random init.
Is this model unavaliable now?
@LuoXin-s you can uncomment https://github.com/huggingface/pytorch-image-models/blob/main/timm/models/eva.py#L820 to download from the original source, the conversion mapping should load it... there is no information as to what that 'pretrain' is, so I've left it disabled for now, the CLIP enormous weights are there.