videomae-finetuned_41

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: 7.9297
• Accuracy: 0.3317

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: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.2
• training_steps: 61640

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.5175	0.05	3083	4.6977	0.3008
0.5311	1.05	6166	5.6012	0.3081
0.5884	2.05	9249	6.2252	0.3175
0.5206	3.05	12332	6.7917	0.3248
0.4449	4.05	15415	6.1943	0.3008
0.3783	5.05	18498	6.8339	0.3150
0.5032	6.05	21581	6.6566	0.3077
0.4091	7.05	24664	6.8013	0.2972
0.4436	8.05	27747	6.8549	0.3
0.3474	9.05	30830	7.0015	0.3268
0.2151	10.05	33913	7.7671	0.3041
0.3597	11.05	36996	7.0724	0.3293
0.1673	12.05	40079	7.5805	0.3248
0.114	13.05	43162	7.8196	0.3175
0.2088	14.05	46245	7.7103	0.3272
0.1662	15.05	49328	7.7613	0.3248
0.1961	16.05	52411	7.7730	0.3297
0.1436	17.05	55494	7.9297	0.3317
0.1134	18.05	58577	8.0447	0.3268
0.0634	19.05	61640	7.9717	0.3305

Framework versions

• Transformers 4.38.1
• Pytorch 2.0.1+cu118
• Datasets 2.18.0
• Tokenizers 0.15.2