MCG-NJU / VideoMAE

[NeurIPS 2022 Spotlight] VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training
https://arxiv.org/abs/2203.12602
Other
1.38k stars 136 forks source link

TypeError: VisionTransformer.__init__() got an unexpected keyword argument 'pretrained_cfg' #126

Open FVT34U opened 3 months ago

FVT34U commented 3 months ago

I have this problem with running code:

Traceback (most recent call last): File "path\to\VideoMAEv2\run_class_finetuning.py", line 926, in main(opts, ds_init) File "path\to\VideoMAEv2\run_class_finetuning.py", line 514, in main model = create_model( File "path\to\anaconda3\envs\videomae\lib\site-packages\timm\models_factory.py", line 117, in create_model model = create_fn( File "path\to\VideoMAEv2\models\modeling_finetune.py", line 459, in vit_small_patch16_224 model = VisionTransformer( TypeError: VisionTransformer.init() got an unexpected keyword argument 'pretrained_cfg'

liuyifan6613 commented 3 months ago

pip install timm==0.4.12