videomae-base-cart-activity-v2-imagefolder

This model is a fine-tuned version of MCG-NJU/videomae-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1807
• Accuracy: 0.9533

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 2
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 17145

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.3978	0.0334	572	0.9201	0.5969
0.294	1.0334	1144	0.3255	0.9027
0.5046	2.0334	1716	0.6951	0.7712
0.3563	3.0334	2288	0.2648	0.9284
0.4592	4.0334	2860	0.2469	0.9167
0.1586	5.0334	3432	0.2294	0.9354
0.1798	6.0334	4004	0.2167	0.9346
0.3834	7.0334	4576	0.2334	0.9292
0.1425	8.0334	5148	0.2864	0.9253
0.1253	9.0334	5720	0.2605	0.9183
0.2152	10.0334	6292	0.2687	0.9276
0.1605	11.0334	6864	0.2657	0.9237
0.1927	12.0334	7436	0.2134	0.9424
0.1003	13.0334	8008	0.2687	0.9300
0.2421	14.0334	8580	0.2345	0.9307
0.1248	15.0334	9152	0.3142	0.9136
0.2701	16.0334	9724	0.2174	0.9409
0.1817	17.0334	10296	0.2328	0.9416
0.1447	18.0334	10868	0.2165	0.9393
0.1101	19.0334	11440	0.2811	0.9300
0.1832	20.0334	12012	0.2230	0.9424
0.1064	21.0334	12584	0.1843	0.9479
0.2109	22.0334	13156	0.2523	0.9323
0.0486	23.0334	13728	0.2485	0.9377
0.1455	24.0334	14300	0.1807	0.9533
0.0695	25.0334	14872	0.2349	0.9393
0.2897	26.0334	15444	0.2229	0.9393
0.0949	27.0334	16016	0.2156	0.9486
0.0849	28.0334	16588	0.2122	0.9479
0.0957	29.0325	17145	0.2072	0.9502

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0