videomae-base-finetuned-engine-subset-20230310

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: 0.4958
• Accuracy: 0.85

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: 6
• eval_batch_size: 6
• 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: 600

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.5947	0.05	31	2.5383	0.15
2.4195	1.05	62	2.5108	0.15
2.2476	2.05	93	2.0533	0.225
1.9449	3.05	124	2.0719	0.2375
1.5724	4.05	155	1.4756	0.475
1.395	5.05	186	1.2884	0.5
1.0822	6.05	217	1.0679	0.575
1.0635	7.05	248	0.8040	0.7
0.8707	8.05	279	0.9334	0.525
0.7042	9.05	310	0.6477	0.75
0.6543	10.05	341	0.6963	0.7375
0.6807	11.05	372	0.4958	0.85
0.4924	12.05	403	0.6374	0.775
0.4822	13.05	434	0.6145	0.75
0.4878	14.05	465	0.6274	0.7625
0.4442	15.05	496	0.4231	0.85
0.2739	16.05	527	0.4999	0.85
0.3514	17.05	558	0.4639	0.8375
0.4158	18.05	589	0.4291	0.85
0.2689	19.02	600	0.4294	0.85

Framework versions

• Transformers 4.26.1
• Pytorch 1.12.1+cu113
• Datasets 2.10.1
• Tokenizers 0.13.2