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: 0.5126
• Accuracy: 0.7586

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: 16
• eval_batch_size: 16
• 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: 150

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.9276	0.0933	14	0.8787	0.6207
0.8845	1.0933	28	0.7380	0.7069
0.685	2.0933	42	0.7084	0.7069
0.7556	3.0933	56	0.6161	0.8103
0.5482	4.0933	70	0.6827	0.7414
0.6461	5.0933	84	0.6916	0.7069
0.701	6.0933	98	0.5490	0.8103
0.5242	7.0933	112	0.5372	0.7931
0.4482	8.0933	126	0.5276	0.8103
0.3898	9.0933	140	0.5047	0.7931
0.3664	10.0667	150	0.5126	0.7586

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.0
• Datasets 3.0.1
• Tokenizers 0.20.0