videomae-large-finetuned-right-hand-conflab-v1

This model is a fine-tuned version of MCG-NJU/videomae-large on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.5240
• Accuracy: 0.6146

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 1404

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0232	0.0420	59	1.9422	0.1942
1.8426	1.0420	118	1.7418	0.3398
1.7424	2.0420	177	1.6896	0.4175
1.2206	3.0420	236	1.6280	0.4466
1.0738	4.0420	295	1.2310	0.5825
1.0054	5.0420	354	1.3243	0.5583
0.782	6.0420	413	1.1891	0.6359
0.599	7.0420	472	1.1930	0.6505
0.6782	8.0420	531	1.2866	0.6359
0.3033	9.0420	590	1.4236	0.5777
0.2236	10.0420	649	1.3206	0.6553
0.1756	11.0420	708	1.5113	0.6602
0.1341	12.0420	767	1.6544	0.6408
0.0823	13.0420	826	1.6124	0.6553
0.0691	14.0420	885	1.8230	0.6456

Framework versions

• Transformers 4.41.0
• Pytorch 1.12.0+cu116
• Datasets 2.19.1
• Tokenizers 0.19.1