videomae-base-finetuned-ucf101-relevancedetection-surgical

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:

• Loss: 0.0913
• Accuracy: 0.9828

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: 7370

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.2187	0.0501	369	2.3587	0.5141
0.2004	1.0501	738	0.3347	0.8992
0.0702	2.0501	1107	0.1048	0.9657
0.0174	3.0501	1476	0.1231	0.9637
0.0463	4.0501	1845	0.5684	0.9012
0.0008	5.0501	2214	1.7179	0.7702
0.022	6.0501	2583	0.4177	0.8972
0.0055	7.0501	2952	0.0563	0.9738
0.0303	8.0501	3321	0.2138	0.9657
0.001	9.0501	3690	0.4224	0.9133
0.0004	10.0501	4059	0.2029	0.9577
0.0003	11.0501	4428	0.4289	0.9375
0.0002	12.0501	4797	0.5915	0.9214
0.0001	13.0501	5166	0.1033	0.9718
0.0001	14.0501	5535	0.1252	0.9677
0.0002	15.0501	5904	0.1371	0.9617
0.0001	16.0501	6273	0.0273	0.9919
0.0002	17.0501	6642	0.0450	0.9899
0.0001	18.0501	7011	0.0284	0.9899
0.0001	19.0487	7370	0.0224	0.9940

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1