videomae-base-finetuned-kinetics-0410_final_5sec_org_ab7_val_inside_train_04

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: 0.3985
• Accuracy: 0.9104

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: 67400

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.0013	0.0100	675	0.5411	0.8395
0.0045	1.0100	1350	0.3895	0.8786
0.0012	2.0100	2025	0.4333	0.8912
0.0011	3.0100	2700	0.3835	0.8998
0.0003	4.0100	3375	0.3007	0.9209
0.0004	5.0100	4050	0.3999	0.8982
0.0003	6.0100	4725	0.3077	0.9303
0.0006	7.0100	5400	0.3264	0.9178
0.0	8.0100	6075	1.3662	0.7643
0.0	9.0100	6750	0.5372	0.8802
0.0	10.0100	7425	0.6177	0.8841
0.0708	11.0100	8100	0.3893	0.9233
0.0005	12.0100	8775	0.5713	0.8865
0.0118	13.0100	9450	0.3995	0.9060
0.0005	14.0100	10125	0.4559	0.8998
0.0001	15.0100	10800	0.6601	0.8904
0.0003	16.0100	11475	0.7191	0.8849
0.0001	17.0100	12150	0.4733	0.9045
0.0122	18.0100	12825	0.4577	0.9154
0.0001	19.0100	13500	0.4770	0.9052
0.0	20.0100	14175	0.4456	0.9178
0.0	21.0100	14850	0.4900	0.9099

Framework versions

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