CTMAE-P2-V2-S1

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: 0.4718
• Accuracy: 0.8261

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: 4
• eval_batch_size: 4
• 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: 3250

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.62	0.02	65	0.7161	0.5435
0.5694	1.02	130	0.7819	0.5435
0.546	2.02	195	0.8927	0.5435
0.6022	3.02	260	0.6859	0.5435
0.5779	4.02	325	0.6449	0.5435
0.4662	5.02	390	0.8167	0.5435
0.5101	6.02	455	0.5114	0.7826
0.3779	7.02	520	0.5149	0.7391
0.3656	8.02	585	0.6273	0.6304
0.4837	9.02	650	0.9093	0.6522
0.6897	10.02	715	0.5653	0.6739
0.435	11.02	780	0.4927	0.7826
0.6362	12.02	845	0.5877	0.6739
0.4422	13.02	910	0.5351	0.8043
0.3913	14.02	975	0.7300	0.8043
0.6191	15.02	1040	1.1917	0.5652
0.2704	16.02	1105	0.5930	0.7826
0.3976	17.02	1170	0.5296	0.8043
0.3038	18.02	1235	0.6735	0.7609
0.2974	19.02	1300	0.4718	0.8261
0.2434	20.02	1365	0.5224	0.8261
0.4984	21.02	1430	1.2637	0.6957
0.1256	22.02	1495	0.7204	0.8261
0.448	23.02	1560	0.6897	0.7609
0.2702	24.02	1625	0.6801	0.8261
0.5101	25.02	1690	0.5134	0.8261
0.354	26.02	1755	0.8076	0.8043
0.4218	27.02	1820	0.7551	0.7826
1.1586	28.02	1885	1.1514	0.6522
0.3586	29.02	1950	1.1479	0.7391
0.4746	30.02	2015	0.9521	0.7174
0.6256	31.02	2080	0.8559	0.8043
0.4668	32.02	2145	0.9766	0.7826
0.1502	33.02	2210	0.9262	0.7826
0.5093	34.02	2275	0.9402	0.7609
0.2621	35.02	2340	0.9229	0.7609
0.1456	36.02	2405	0.7937	0.8261
0.1826	37.02	2470	0.9106	0.7826
0.3778	38.02	2535	0.9376	0.7826
0.1763	39.02	2600	0.9300	0.7826
0.1083	40.02	2665	1.1018	0.7609
0.1994	41.02	2730	0.8667	0.8261
0.0111	42.02	2795	0.9896	0.8043
0.0818	43.02	2860	1.0258	0.7826
0.1808	44.02	2925	0.9841	0.7826
0.1371	45.02	2990	0.9337	0.8043
0.0129	46.02	3055	0.8905	0.8043
0.1492	47.02	3120	0.9629	0.8261
0.0184	48.02	3185	1.0828	0.7174
0.1146	49.02	3250	1.0449	0.7826

Framework versions

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