finetuned-for-YogaPosesv6

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the yoga_pose_images dataset. It achieves the following results on the evaluation set:

• Loss: 0.0562
• Accuracy: 0.9938

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: 0.0002
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• 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_steps: 500
• num_epochs: 12
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.9954	0.8772	100	0.8301	0.8505
0.3505	1.7544	200	0.1881	0.9907
0.1524	2.6316	300	0.0901	0.9844
0.152	3.5088	400	0.1241	0.9688
0.1314	4.3860	500	0.0562	0.9938
0.1187	5.2632	600	0.1232	0.9720
0.0936	6.1404	700	0.0893	0.9688
0.0753	7.0175	800	0.1510	0.9626
0.0155	7.8947	900	0.0536	0.9907
0.0181	8.7719	1000	0.0515	0.9907
0.0037	9.6491	1100	0.0570	0.9907
0.0121	10.5263	1200	0.0570	0.9907
0.0065	11.4035	1300	0.0565	0.9907

Framework versions

• Transformers 4.46.3
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3