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: 0.0794
• Accuracy: 0.9714

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
2.1779	0.0833	50	2.0389	0.2714
0.9209	1.0833	100	0.9262	0.6857
0.5527	2.0833	150	0.3633	0.9143
0.2367	3.0833	200	0.4540	0.8857
0.4635	4.0833	250	0.2192	0.9429
0.097	5.0833	300	0.2792	0.8714
0.0128	6.0833	350	0.1230	0.9571
0.0346	7.0833	400	0.0637	0.9714
0.005	8.0833	450	0.0655	0.9714
0.0045	9.0833	500	0.0876	0.9714
0.004	10.0833	550	0.0904	0.9714
0.0041	11.0833	600	0.0794	0.9714

Framework versions

• Transformers 4.45.1
• Pytorch 1.13.1+cu117
• Datasets 3.0.1
• Tokenizers 0.20.0