ViViT_LSA64_SR_6

This model is a fine-tuned version of google/vivit-b-16x2-kinetics400 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0190
• Accuracy: 0.9961
• Precision: 0.9969
• Recall: 0.9961
• F1: 0.9960

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: 8640
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
15.7817	0.0333	288	2.8824	0.4609	0.5309	0.4609	0.4307
4.7558	1.0333	576	0.5582	0.9492	0.9586	0.9492	0.9470
0.5173	2.0333	864	0.0999	0.9805	0.9854	0.9805	0.9798
0.1244	3.0333	1152	0.0102	1.0	1.0	1.0	1.0
0.0043	4.0333	1440	0.0265	0.9922	0.9938	0.9922	0.9921
0.021	5.0333	1728	0.0200	0.9922	0.9938	0.9922	0.9921
0.0014	6.0333	2016	0.0012	1.0	1.0	1.0	1.0
0.0414	7.0333	2304	0.0075	0.9961	0.9969	0.9961	0.9960
0.0386	8.0333	2592	0.0190	0.9961	0.9969	0.9961	0.9960

Framework versions

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