test_model_88

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the corranm/first_vote_100_per_new2 dataset. It achieves the following results on the evaluation set:

• Loss: 1.8934
• F1 Macro: 0.0606
• F1 Micro: 0.1591
• F1 Weighted: 0.0846
• Precision Macro: 0.0421
• Precision Micro: 0.1591
• Precision Weighted: 0.0586
• Recall Macro: 0.1132
• Recall Micro: 0.1591
• Recall Weighted: 0.1591
• Accuracy: 0.1591

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• 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: 3

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Macro	F1 Micro	F1 Weighted	Precision Macro	Precision Micro	Precision Weighted	Recall Macro	Recall Micro	Recall Weighted	Accuracy
1.9682	0.8	3	1.9070	0.0599	0.2121	0.0848	0.0661	0.2121	0.0908	0.1486	0.2121	0.2121	0.2121
1.8993	1.8	6	1.8860	0.0902	0.2197	0.1243	0.0630	0.2197	0.0867	0.1594	0.2197	0.2197	0.2197
2.3539	2.8	9	1.8915	0.0637	0.1591	0.0887	0.0443	0.1591	0.0616	0.1141	0.1591	0.1591	0.1591

Framework versions

• Transformers 4.48.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0