videomae-base-finetuned-kinetics-finetuned-right-hand-conflab-baseline-2

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: 1.4525
• Accuracy: 0.5951

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: 1170

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.7757	0.0504	59	1.7544	0.3398
1.4807	1.0504	118	1.7067	0.3689
1.4444	2.0504	177	1.4684	0.4320
1.022	3.0504	236	1.4322	0.5388
0.7796	4.0504	295	1.2440	0.5583
0.5612	5.0504	354	1.3034	0.5680
0.4722	6.0504	413	1.4143	0.5728
0.3215	7.0504	472	1.3449	0.5825
0.2223	8.0504	531	1.3282	0.6019
0.1452	9.0504	590	1.4134	0.6068
0.1771	10.0504	649	1.4489	0.6019
0.0472	11.0504	708	1.5544	0.5922
0.0384	12.0504	767	1.6467	0.5874

Framework versions

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