9_mae_3

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: 0.6876
• Accuracy: 0.7609

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: 9750

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7285	0.02	195	0.7801	0.4565
0.6181	1.02	390	0.6808	0.4565
0.4805	2.02	585	0.5866	0.7174
0.6863	3.02	780	0.7855	0.5870
0.4042	4.02	975	0.6876	0.7609
0.6607	5.02	1170	0.6689	0.4565
0.4924	6.02	1365	0.7538	0.5217
0.7448	7.02	1560	0.6035	0.7391
0.3786	8.02	1755	1.6179	0.4565
0.5396	9.02	1950	0.9411	0.6304
0.4499	10.02	2145	1.1186	0.6087
0.4792	11.02	2340	1.1845	0.5652
0.3171	12.02	2535	0.7825	0.7609
0.5184	13.02	2730	0.8138	0.6522
0.3237	14.02	2925	0.7518	0.7391
0.5328	15.02	3120	0.6445	0.7391
0.2974	16.02	3315	1.1197	0.7609
0.9279	17.02	3510	0.6909	0.7609
0.3876	18.02	3705	1.0378	0.6087
0.3241	19.02	3900	1.2621	0.6739
0.2565	20.02	4095	1.2284	0.6957
0.4965	21.02	4290	1.4872	0.6304
0.2334	22.02	4485	1.5145	0.6739
0.1407	23.02	4680	0.9717	0.7609
0.493	24.02	4875	1.3058	0.7609
0.1002	25.02	5070	1.1231	0.7609
0.2103	26.02	5265	1.4007	0.7174
0.1773	27.02	5460	1.5471	0.7174
0.5669	28.02	5655	1.8812	0.6739
0.092	29.02	5850	1.3838	0.7609
0.136	30.02	6045	1.9547	0.6957
0.2221	31.02	6240	1.9051	0.6957
0.3002	32.02	6435	1.9346	0.6957
0.3202	33.02	6630	1.9951	0.6304
0.004	34.02	6825	1.9543	0.6304
0.0914	35.02	7020	2.0569	0.6739
0.0853	36.02	7215	1.7665	0.7174
0.2405	37.02	7410	2.0684	0.6739
0.0649	38.02	7605	2.1542	0.6957
0.0128	39.02	7800	1.8697	0.6739
0.353	40.02	7995	1.7250	0.6957
0.0011	41.02	8190	1.7851	0.6957
0.0007	42.02	8385	1.7935	0.6739
0.3829	43.02	8580	1.8022	0.6957
0.0004	44.02	8775	1.8407	0.6739
0.1538	45.02	8970	1.6977	0.7391
0.1893	46.02	9165	2.0842	0.7174
0.3183	47.02	9360	1.8477	0.7174
0.1938	48.02	9555	2.0269	0.7174
0.123	49.02	9750	2.0875	0.7174

Framework versions

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