videomae-base-finetuned-ucf101-subset

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: 0.1719
• Accuracy: 0.9548

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: 8
• eval_batch_size: 8
• 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
2.2101	0.0633	38	2.1049	0.3143
1.286	1.0633	76	1.1230	0.5286
0.4989	2.0633	114	0.4172	0.8714
0.4114	3.0633	152	0.3117	0.8286
0.1907	4.0633	190	0.2904	0.9
0.1088	5.0633	228	0.1609	0.9571
0.057	6.0633	266	0.0535	0.9857
0.117	7.0633	304	0.2749	0.9143
0.1324	8.0633	342	0.1784	0.9429
0.0081	9.0633	380	0.0184	1.0
0.008	10.0633	418	0.0355	0.9857
0.0038	11.0633	456	0.0087	1.0
0.0033	12.0633	494	0.0086	1.0
0.0033	13.0633	532	0.0506	0.9857
0.0035	14.0633	570	0.0357	0.9857
0.0031	15.05	600	0.0365	0.9857

Framework versions

• Transformers 4.41.2
• Pytorch 2.0.0+cu117
• Datasets 2.19.2
• Tokenizers 0.19.1