videomae-base-finetuned-kinetics-allkisa-crop-background-0228

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: 0.3129
• Accuracy: 0.9163

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: 10560

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.0035	0.0251	265	0.7057	0.7302
0.0047	1.0251	530	0.6369	0.7840
0.3625	2.0251	795	1.0777	0.7099
0.017	3.0251	1060	1.1492	0.6145
0.0001	4.0251	1325	0.5946	0.8394
0.0006	5.0251	1590	0.4763	0.8134
0.0036	6.0251	1855	0.4798	0.8093
0.0286	7.0251	2120	0.6486	0.8191
0.0003	8.0251	2385	0.8464	0.7775
0.0001	9.0251	2650	0.6891	0.8174
0.0003	10.0251	2915	0.5298	0.8582
0.0003	11.0251	3180	0.7832	0.8289
0.0	12.0251	3445	0.7445	0.8215
0.0	13.0251	3710	0.6594	0.8419
0.0	14.0251	3975	0.7089	0.8297
0.0001	15.0251	4240	0.6542	0.8509
0.0	16.0251	4505	0.7677	0.8346
0.0	17.0251	4770	0.8255	0.8183
0.0011	18.0251	5035	0.7238	0.8378
0.0001	19.0251	5300	0.7826	0.8329
0.0001	20.0251	5565	0.7980	0.8329

Framework versions

• Transformers 4.48.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0