MAE-CT-CPC-Dicotomized-v4-early-stop

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

• Loss: 0.4282
• Accuracy: 0.8293

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: 4
• eval_batch_size: 4
• 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: 1440

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.5342	0.06	81	0.6496	0.6829
0.6266	1.06	162	0.5854	0.6829
0.6599	2.06	243	0.5206	0.6829
0.877	3.06	324	0.5995	0.6098
0.653	4.06	405	0.4908	0.7561
0.7604	5.06	486	0.4936	0.7805
0.4795	6.06	567	0.9528	0.6829
0.278	7.06	648	0.5565	0.8049
0.3548	8.06	729	0.5855	0.7561
0.4386	9.06	810	0.6578	0.7561
0.3007	10.06	891	0.6622	0.7805
0.313	11.06	972	0.8350	0.7561
0.0554	12.06	1053	1.0043	0.7073
0.2804	13.06	1134	1.0247	0.7073
0.1424	14.06	1215	0.8542	0.7805
0.4692	15.06	1296	1.0264	0.7317

Framework versions

• Transformers 4.37.2
• Pytorch 2.0.1+cu117
• Datasets 3.0.1
• Tokenizers 0.15.1