videomae-base-finetuned-isl-numbers-alphabet-nouns

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: 0.4278
• Accuracy: 0.8875

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: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 15800

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
4.5228	0.02	316	4.3514	0.2534
3.0795	1.02	632	2.8515	0.5816
1.8438	2.02	948	1.7508	0.7332
1.1451	3.02	1264	1.1464	0.7390
1.0637	4.02	1580	0.7995	0.7774
0.7795	5.02	1896	0.4938	0.8829
0.4484	6.02	2212	0.3833	0.8829
0.2162	7.02	2528	0.2512	0.9155
0.228	8.02	2844	0.1972	0.9309
0.1711	9.02	3160	0.1426	0.9482
0.2251	10.02	3476	0.0965	0.9559
0.1697	11.02	3792	0.1141	0.9539
0.1229	12.02	4108	0.1362	0.9539
0.0676	13.02	4424	0.0745	0.9655
0.1228	14.02	4740	0.0817	0.9635
0.0143	15.02	5056	0.0615	0.9693
0.0621	16.02	5372	0.0768	0.9597
0.0597	17.02	5688	0.0873	0.9635
0.0696	18.02	6004	0.1108	0.9539
0.2761	19.02	6320	0.1413	0.9520
0.129	20.02	6636	0.1471	0.9520
0.0828	21.02	6952	0.0608	0.9674
0.0544	22.02	7268	0.0533	0.9712
0.0509	23.02	7584	0.0499	0.9750
0.0308	24.02	7900	0.0956	0.9597
0.0729	25.02	8216	0.0753	0.9731
0.2328	26.02	8532	0.0774	0.9655
0.1085	27.02	8848	0.0609	0.9693
0.099	28.02	9164	0.0677	0.9674
0.1988	29.02	9480	0.1415	0.9559
0.0747	30.02	9796	0.0581	0.9712
0.0556	31.02	10112	0.0519	0.9693
0.0763	32.02	10428	0.0506	0.9731
0.0635	33.02	10744	0.0492	0.9750
0.0729	34.02	11060	0.0483	0.9693
0.0692	35.02	11376	0.0481	0.9750
0.1023	36.02	11692	0.0478	0.9712
0.0863	37.02	12008	0.0479	0.9750
0.0934	38.02	12324	0.0464	0.9712
0.0927	39.02	12640	0.0462	0.9712
0.0254	40.02	12956	0.0448	0.9731
0.043	41.02	13272	0.0450	0.9750
0.0695	42.02	13588	0.0448	0.9750
0.0398	43.02	13904	0.0440	0.9770
0.0455	44.02	14220	0.0436	0.9770
0.0423	45.02	14536	0.0437	0.9750
0.0602	46.02	14852	0.0438	0.9770
0.0407	47.02	15168	0.0437	0.9750
0.0435	48.02	15484	0.0435	0.9770
0.0463	49.02	15800	0.0436	0.9770

Framework versions

• Transformers 4.40.0
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.19.1