videomae-base-finetuned-ucfcrime-full2

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: 2.5470
• Accuracy: 0.2143

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: 800
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.2039	0.125	100	2.4597	0.1644
1.8182	1.125	200	2.3359	0.2046
1.9124	2.125	300	2.6805	0.2078
1.7017	3.125	400	2.3331	0.2453
1.68	4.125	500	2.2637	0.2225
1.5457	5.125	600	2.4711	0.2284
1.4727	6.125	700	2.3691	0.2110
1.5393	7.125	800	2.4561	0.2025

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0
• Datasets 3.0.0
• Tokenizers 0.19.1