videomae-base-finetuned-ucf101-subset

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: 2.1806
• Accuracy: 0.4883

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: 32
• eval_batch_size: 32
• 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: 960

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
4.2427	0.0323	31	4.2265	0.0033
4.2321	1.0323	62	4.2235	0.0100
4.24	2.0323	93	4.2282	0.0100
4.2445	3.0323	124	4.2250	0.0067
4.2327	4.0323	155	4.2244	0.0100
4.2104	5.0323	186	4.2100	0.0201
4.2374	6.0323	217	4.2022	0.0067
4.1597	7.0323	248	4.1188	0.0301
4.0522	8.0323	279	3.9351	0.0702
3.768	9.0323	310	3.6800	0.1070
3.5147	10.0323	341	3.5416	0.1104
3.2878	11.0323	372	3.7074	0.0702
2.9491	12.0323	403	3.3954	0.1070
2.806	13.0323	434	3.2552	0.1706
2.4568	14.0323	465	3.0654	0.2040
2.3102	15.0323	496	2.7440	0.3010
2.2079	16.0323	527	2.6789	0.3144
1.9638	17.0323	558	2.5920	0.3679
1.7914	18.0323	589	2.6152	0.3378
1.6925	19.0323	620	2.5971	0.3445
1.5124	20.0323	651	2.5767	0.3478
1.4834	21.0323	682	2.4439	0.3880
1.4565	22.0323	713	2.4057	0.3846
1.279	23.0323	744	2.5501	0.3545
1.1477	24.0323	775	2.3247	0.4482
1.2573	25.0323	806	2.1776	0.4883
1.0825	26.0323	837	2.1443	0.4783
1.2121	27.0323	868	2.1490	0.4783
1.0887	28.0323	899	2.1516	0.4716
1.1127	29.0323	930	2.1051	0.4883
0.9905	30.0312	960	2.1170	0.4816

Framework versions

• Transformers 4.48.0
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0