ALL_RGBCROP_ori16F-8B16F-GACWDlr5e-6

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: 0.4757
• Accuracy: 0.8443

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-06
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.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: 768

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6952	0.0625	48	0.7022	0.5244
0.5982	1.0625	96	0.6475	0.6463
0.4329	2.0625	144	0.5911	0.7134
0.3752	3.0625	192	0.5338	0.7378
0.2202	4.0625	240	0.5109	0.75
0.1893	5.0625	288	0.4991	0.7439
0.1454	6.0625	336	0.5313	0.7439
0.0931	7.0625	384	0.5668	0.7439
0.0657	8.0625	432	0.6028	0.7622
0.0704	9.0625	480	0.6340	0.7744
0.0298	10.0625	528	0.6770	0.7683
0.031	11.0625	576	0.7089	0.7805
0.0265	12.0625	624	0.7415	0.7683
0.0238	13.0625	672	0.7703	0.7744
0.0143	14.0625	720	0.7857	0.7744
0.0082	15.0625	768	0.7922	0.7683

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1