videomae-base-finetuned-kinetics-0404_final_45sec

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

• Loss: 1.1708
• Accuracy: 0.7595

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: 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: 0.05
• training_steps: 63000

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.4251	0.0100	631	0.7287	0.8076
0.5829	1.0100	1262	0.7866	0.7696
0.3161	2.0100	1893	0.4666	0.8191
0.1667	3.0100	2524	1.3251	0.7493
0.0019	4.0100	3155	1.3841	0.7485
0.006	5.0100	3786	0.7565	0.8358
0.0088	6.0100	4417	1.3582	0.7767
0.0026	7.0100	5048	1.2738	0.7652
0.0015	8.0100	5679	1.0358	0.8173
0.24	9.0100	6310	0.9473	0.8350
0.0065	10.0100	6941	1.2006	0.7696
0.0004	11.0100	7572	1.2525	0.7891
0.0003	12.0100	8203	2.1669	0.7017
0.0004	13.0100	8834	1.5847	0.7696
0.0053	14.0100	9465	0.8446	0.8420
0.0017	15.0100	10096	1.2059	0.8076
0.0019	16.0100	10727	1.6806	0.7432
0.0028	17.0100	11358	1.1255	0.8138
0.737	18.0100	11989	1.6861	0.7626
0.0001	19.0100	12620	1.2263	0.8226
0.0012	20.0100	13251	1.7637	0.7555
0.1868	21.0100	13882	1.5043	0.7988
0.0003	22.0100	14513	1.7795	0.7838
0.0033	23.0100	15144	1.3303	0.8147
0.1839	24.0100	15775	1.5371	0.7882
0.0019	25.0100	16406	1.6585	0.7811
0.0001	26.0100	17037	1.2823	0.8111
0.0151	27.0100	17668	1.1360	0.8376
0.0001	28.0100	18299	1.4864	0.8041
0.0008	29.0100	18930	1.5519	0.7988

Framework versions

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