videomae-base-finetuned-kinetics-final-contest-baole4-0705

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: 0.3771
• Accuracy: 0.8807

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: 4
• eval_batch_size: 4
• 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: 2805

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.4477	0.0670	188	0.8428	0.8486
0.3123	1.0670	376	0.6401	0.8532
0.2498	2.0670	564	0.5204	0.8578
0.0348	3.0670	752	0.4130	0.8899
0.146	4.0670	940	0.3617	0.8991
0.038	5.0670	1128	0.4519	0.8670
0.0116	6.0670	1316	0.4160	0.8716
0.0083	7.0670	1504	0.3683	0.8807
0.0067	8.0670	1692	0.3820	0.8807
0.0051	9.0670	1880	0.3932	0.8761
0.0057	10.0670	2068	0.3944	0.8853
0.0045	11.0670	2256	0.4133	0.8761
0.0042	12.0670	2444	0.3813	0.8807
0.0037	13.0670	2632	0.3806	0.8853
0.0039	14.0617	2805	0.3771	0.8807

Framework versions

• Transformers 4.40.2
• Pytorch 1.13.1+cu117
• Datasets 2.18.0
• Tokenizers 0.19.1