ViViT_wlasl_100__signer_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: 1.2104
• Accuracy: 0.6864

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: 3600
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
18.6783	0.05	180	4.5403	0.0414
16.851	1.0499	360	3.8917	0.1568
12.3649	2.0499	540	3.0026	0.3284
7.877	3.0501	721	2.3964	0.4852
4.8527	4.05	901	2.0190	0.5651
2.7599	5.0499	1081	1.7904	0.5976
1.516	6.0499	1261	1.5763	0.6272
0.8343	7.0501	1442	1.4678	0.6154
0.4144	8.05	1622	1.4205	0.6420
0.2624	9.0499	1802	1.2896	0.6716
0.1498	10.0499	1982	1.2734	0.6627
0.0906	11.0501	2163	1.2307	0.6805
0.0779	12.05	2343	1.2818	0.6805
0.0647	13.0499	2523	1.2184	0.6893
0.0615	14.0499	2703	1.2204	0.6953
0.0513	15.0501	2884	1.2155	0.6716
0.0456	16.05	3064	1.2052	0.6982
0.0219	17.0499	3244	1.2122	0.6864
0.033	18.0499	3424	1.2159	0.6893
0.0206	19.0487	3600	1.2104	0.6864

Framework versions

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