CTMAE2_CS_V7_2

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

• Loss: 0.7423
• Accuracy: 0.7778

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: 1e-05
• train_batch_size: 3
• eval_batch_size: 3
• 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: 12950

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.5325	0.0201	260	0.8741	0.4667
0.5015	1.0201	520	0.9818	0.4444
0.5515	2.0201	780	0.7134	0.5111
0.3752	3.0201	1040	0.7412	0.6222
0.3128	4.0201	1300	0.8534	0.6
0.7261	5.0201	1560	0.7002	0.7111
0.4644	6.0201	1820	0.6550	0.7111
0.7702	7.0201	2080	0.6853	0.7333
0.4589	8.0201	2340	0.7447	0.7556
0.5026	9.0201	2600	0.7423	0.7778
0.5612	10.0201	2860	0.8798	0.6889
0.536	11.0201	3120	1.5358	0.4889
0.6695	12.0201	3380	1.3352	0.6667
0.2699	13.0201	3640	1.1053	0.7333
0.5277	14.0201	3900	0.9908	0.7111
0.7975	15.0201	4160	1.0846	0.6667
0.5766	16.0201	4420	0.9612	0.7333
0.3323	17.0201	4680	1.1611	0.6889
0.7162	18.0201	4940	1.3055	0.7111
0.1248	19.0201	5200	1.5837	0.6667
0.1746	20.0201	5460	1.2694	0.7333
0.1973	21.0201	5720	1.1572	0.7778
0.7038	22.0201	5980	1.4035	0.7111
0.3939	23.0201	6240	1.6775	0.6667
0.4129	24.0201	6500	1.6039	0.6889
0.2852	25.0201	6760	1.6769	0.7111
0.3278	26.0201	7020	1.9129	0.6889
0.7677	27.0201	7280	1.8397	0.6667
0.0122	28.0201	7540	2.0296	0.6889
0.3014	29.0201	7800	2.4234	0.6
0.3567	30.0201	8060	1.7570	0.7111
0.0334	31.0201	8320	2.0343	0.7111
0.0043	32.0201	8580	1.8095	0.7333
0.0119	33.0201	8840	1.6490	0.7556
0.7503	34.0201	9100	1.9144	0.6889
0.2105	35.0201	9360	1.8403	0.7333
0.003	36.0201	9620	1.8770	0.7333
0.1781	37.0201	9880	1.8631	0.7333
0.4092	38.0201	10140	1.9994	0.7111
0.232	39.0201	10400	1.9919	0.6889
0.2703	40.0201	10660	2.1008	0.7111
0.5169	41.0201	10920	2.2019	0.6889
0.8418	42.0201	11180	2.2000	0.6889
0.0007	43.0201	11440	2.0411	0.7111
0.0001	44.0201	11700	2.1081	0.6889
0.0004	45.0201	11960	2.1821	0.6889
0.0013	46.0201	12220	2.1313	0.7111
0.0003	47.0201	12480	2.2113	0.7111
0.83	48.0201	12740	2.2048	0.6889
0.0001	49.0162	12950	2.1997	0.6889

Framework versions

• Transformers 4.46.2
• Pytorch 2.0.1+cu117
• Datasets 3.0.1
• Tokenizers 0.20.0