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: 0.1199
• 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: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• 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
2.2409	0.06	37	2.2371	0.1429
1.3169	1.06	75	1.1241	0.6571
0.5831	2.06	112	0.5958	0.7857
0.5517	3.06	150	0.4112	0.8143
0.398	4.06	187	0.3376	0.8429
0.1959	5.06	225	0.4228	0.8857
0.1159	6.06	262	0.3382	0.8571
0.015	7.06	300	0.3205	0.9
0.0316	8.06	337	0.3495	0.8857
0.0242	9.06	375	0.1675	0.9429
0.005	10.06	412	0.2990	0.9286
0.0047	11.06	450	0.1553	0.9429
0.0044	12.06	487	0.1390	0.9571
0.0039	13.06	525	0.1406	0.9429
0.0107	14.06	562	0.1184	0.9571
0.0034	15.06	600	0.1199	0.9714

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.0
• Tokenizers 0.13.3