VideoMAE_BdSLW60_FrameRate_Corrected_WITHOUT_Augment_20_epoch_RQ_GB

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: 1.5110
• Accuracy: 0.75
• Precision: 0.7751
• Recall: 0.75
• F1: 0.7160

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	Precision	Recall	F1
9.37	0.0501	929	2.6435	0.3867	0.3678	0.3867	0.3116
1.0137	1.0501	1858	1.1730	0.6883	0.7046	0.6883	0.6376
0.2085	2.0501	2787	0.7637	0.7817	0.7970	0.7817	0.7487
0.1241	3.0501	3716	1.2059	0.7383	0.7799	0.7383	0.7021
0.0466	4.0501	4645	1.5994	0.7317	0.7389	0.7317	0.6956
0.2235	5.0501	5574	0.8861	0.815	0.8428	0.815	0.7941
0.1335	6.0501	6503	1.6215	0.75	0.7610	0.75	0.7153
0.0882	7.0501	7432	1.5239	0.7583	0.7890	0.7583	0.7463
0.0716	8.0501	8361	1.4319	0.775	0.7999	0.775	0.7471
0.0076	9.0501	9290	1.7182	0.7533	0.7603	0.7533	0.7160
0.2277	10.0501	10219	1.5110	0.75	0.7751	0.75	0.7160

Framework versions

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