VideoMAE_kINETICS_BDSLW60_U5_20EPOCH

This model is a fine-tuned version of MCG-NJU/videomae-base-finetuned-kinetics on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2234
• Accuracy: 0.955

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 18560
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
10.3766	0.0501	929	0.8142	0.905
0.7747	1.0501	1858	0.1567	0.96
0.158	2.0501	2787	0.1637	0.965
0.0971	3.0501	3716	0.1483	0.9633
0.0827	4.0501	4645	0.1992	0.9567
0.1003	5.0501	5574	0.1753	0.9567
0.079	6.0501	6503	0.1162	0.97
0.0835	7.0501	7432	0.0722	0.9817
0.0943	8.0501	8361	0.2166	0.9617
0.0532	9.0501	9290	0.2413	0.9633
0.0618	10.0501	10219	0.1646	0.965
0.0349	11.0501	11148	0.2529	0.955
0.0384	12.0501	12077	0.2234	0.955

Framework versions

• Transformers 4.46.1
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.20.1