videomae-base-finetuned-good-gesturePhaseV11

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

• Accuracy: 0.9461
• Loss: 0.1967
• Accuracy Hold: 1.0
• Accuracy Stroke: 0.4286
• Accuracy Recovery: 0.8947
• Accuracy Preparation: 0.9686
• Accuracy Unknown: 0.9286

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• 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: 1260

Training results

Training Loss	Epoch	Step	Accuracy	Validation Loss	Accuracy Hold	Accuracy Stroke	Accuracy Recovery	Accuracy Preparation	Accuracy Unknown
1.2097	0.1008	127	0.6900	1.0243	0.0	0.0	0.0	1.0	0.0
0.8	1.1008	254	0.6900	0.8717	0.0	0.0	0.0	1.0	0.0
0.5199	2.1008	381	0.7729	0.6725	0.0	0.0	0.1765	0.9810	0.8636
0.3134	3.1008	508	0.8428	0.4715	0.3077	0.0	0.4118	1.0	0.9091
0.1561	4.1008	635	0.8952	0.4363	0.7692	0.0	0.7059	1.0	0.6818
0.0429	5.1008	762	0.9432	0.2211	0.8846	0.5	0.7059	0.9937	0.9545
0.2294	6.1008	889	0.9476	0.2094	0.8846	0.1667	0.8824	1.0	0.9091
0.1214	7.1008	1016	0.9607	0.1586	0.8846	0.6667	0.8824	0.9937	0.9545
0.1478	8.1008	1143	0.9432	0.1607	0.8846	0.6667	0.8824	0.9684	0.9545
0.1156	9.0929	1260	0.9738	0.1177	1.0	0.6667	0.8824	0.9937	0.9545

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1