videomae-base-finetuned-kinetics-0417_final_5sec_org_ab7_val_inside_train_balanced_allpretest

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: 0.2601
• Accuracy: 0.9217

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: 86000

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.0156	0.0100	861	0.1954	0.9281
2.8163	1.0100	1722	0.1982	0.9384
2.3325	2.0100	2583	0.4539	0.8791
0.0072	3.0100	3444	0.4816	0.8752
2.4587	4.0100	4305	0.2090	0.9423
0.0005	5.0100	5166	0.2878	0.9139
2.7948	6.0100	6027	0.7033	0.8507
0.0471	7.0100	6888	0.3067	0.9084
0.003	8.0100	7749	0.4804	0.8949
0.0026	9.0100	8610	0.3477	0.9226
0.0007	10.0100	9471	0.3830	0.9194
0.0014	11.0100	10332	0.2867	0.9171
0.0002	12.0100	11193	0.5321	0.8965
0.0016	13.0100	12054	0.3969	0.9100
0.0077	14.0100	12915	0.3807	0.9297
0.0083	15.0100	13776	0.3125	0.9305
0.001	16.0100	14637	0.3918	0.9202
0.0003	17.0100	15498	0.3189	0.9376
0.0001	18.0100	16359	0.3230	0.9360
0.0204	19.0100	17220	0.3610	0.9376

Framework versions

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