videomae-base-finetuned-kinetics-0326_final_imgsz256

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: 1.0012
• Accuracy: 0.6828

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: 68700

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.2669	0.01	687	0.7324	0.7318
0.311	1.01	1374	1.2855	0.6386
0.1627	2.01	2061	0.7297	0.7788
0.4673	3.01	2748	1.7532	0.6191
0.9694	4.01	3435	0.9305	0.7715
0.2372	5.01	4122	1.0085	0.7423
0.3608	6.01	4809	0.7486	0.7618
0.8374	7.01	5496	1.1176	0.7366
0.0445	8.01	6183	1.2131	0.7382
0.6795	9.01	6870	1.7516	0.6345
0.2234	10.01	7557	2.2503	0.5989
0.0306	11.01	8244	1.5213	0.7115
0.1396	12.01	8931	1.3085	0.7253

Framework versions

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