ViViT_wlasl_2000_20ep_coR

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: 3.2256
• Accuracy: 0.3437

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: 35720
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
30.7198	0.05	1786	7.5765	0.0023
29.0292	1.0500	3572	6.7339	0.0255
24.0941	2.0500	5358	5.7016	0.0889
18.973	3.0500	7145	4.8953	0.1604
14.3207	4.05	8931	4.2779	0.2188
10.2288	5.0500	10717	3.8242	0.2561
6.8987	6.0500	12503	3.4816	0.2990
4.4195	7.0500	14290	3.3354	0.3105
2.8005	8.05	16076	3.2289	0.3212
1.8191	9.0500	17862	3.1795	0.3200
1.2778	10.0500	19648	3.1637	0.3292
1.0009	11.0500	21435	3.1523	0.3299
0.8082	12.05	23221	3.1508	0.3292
0.7047	13.0500	25007	3.1626	0.3276
0.6152	14.0500	26793	3.1711	0.3327
0.545	15.0500	28580	3.2040	0.3394
0.4952	16.05	30366	3.1936	0.3424
0.4463	17.0500	32152	3.2133	0.3435
0.403	18.0500	33938	3.2240	0.3432
0.3506	19.0499	35720	3.2256	0.3437

Framework versions

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