food_classifier_2025_01_31_00_04

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

• Loss: 0.4920
• Accuracy: 0.8763

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: 0.0005
• train_batch_size: 128
• eval_batch_size: 128
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 2048
• total_eval_batch_size: 512
• optimizer: Use OptimizerNames.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
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
3.9401	1.0	37	3.1519	0.7044
1.5951	2.0	74	1.1581	0.7973
0.916	3.0	111	0.7583	0.8228
0.7189	4.0	148	0.6624	0.8371
0.5926	5.0	185	0.6070	0.8476
0.5456	6.0	222	0.5709	0.8553
0.4675	7.0	259	0.5564	0.8572
0.4246	8.0	296	0.5465	0.8602
0.3732	9.0	333	0.5401	0.8627
0.333	10.0	370	0.5197	0.8671
0.3067	11.0	407	0.5077	0.8712
0.2872	12.0	444	0.5090	0.8702
0.2537	13.0	481	0.5066	0.8761
0.2496	14.0	518	0.5004	0.8750
0.2282	15.0	555	0.4920	0.8763

Framework versions

• Transformers 4.48.1
• Pytorch 2.5.1
• Datasets 2.19.1
• Tokenizers 0.21.0