ALL_RGBCROP_Aug16F-8B16F-GACWDlr

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: 0.5910
• Accuracy: 0.8743

Best Checkpoint : 1296

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: 1e-05
• 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: 2304

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.5269	0.0625	144	0.5941	0.6646
0.2405	1.0625	288	0.5302	0.7724
0.0421	2.0625	432	0.7070	0.7886
0.006	3.0625	576	0.9248	0.7825
0.0039	4.0625	720	1.0677	0.7907
0.0003	5.0625	864	1.0671	0.8008
0.0002	6.0625	1008	1.1264	0.7967
0.0002	7.0625	1152	1.1610	0.7927
0.0002	8.0625	1296	1.1490	0.8028
0.0001	9.0625	1440	1.1798	0.7967
0.0001	10.0625	1584	1.2032	0.7988
0.0001	11.0625	1728	1.2225	0.7967
0.0001	12.0625	1872	1.2358	0.7988
0.0001	13.0625	2016	1.2433	0.8008

Framework versions

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