videomae-base-finetuned-kinetics-0416_final_5sec_org_ab7_val_inside_train_balanced

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: 0.3263
• Accuracy: 0.9260

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: 62900

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7654	0.0100	630	0.6926	0.7972
0.0117	1.0100	1260	0.3749	0.9052
0.4759	2.0100	1890	0.4432	0.8951
0.9896	3.0100	2520	0.2561	0.9256
1.0067	4.0100	3150	0.6378	0.8074
0.021	5.0100	3780	0.8654	0.8066
0.0797	6.0100	4410	0.3623	0.9146
0.0025	7.0100	5040	0.2818	0.9162
0.7118	8.0100	5670	0.4359	0.8990
0.0003	9.0100	6300	0.2764	0.9327
0.056	10.0100	6930	0.4070	0.9076
0.8554	11.0100	7560	0.2826	0.9381
0.0058	12.0100	8190	0.5072	0.8966
0.0002	13.0100	8820	0.9202	0.8520
0.0038	14.0100	9450	0.3547	0.9295
0.0003	15.0100	10080	0.4670	0.9115
0.0001	16.0100	10710	0.4114	0.9248
0.0059	17.0100	11340	0.6451	0.8958
0.0003	18.0100	11970	0.4430	0.9139
0.0003	19.0100	12600	0.4658	0.9186
0.0043	20.0100	13230	0.5688	0.9021
0.0004	21.0100	13860	0.4696	0.9170
0.0001	22.0100	14490	0.4937	0.9201
0.0008	23.0100	15120	0.6324	0.9021
0.0001	24.0100	15750	0.6696	0.8951
0.9098	25.0100	16380	0.4920	0.9264
0.0006	26.0100	17010	0.5209	0.9225

Framework versions

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