fetal_planes_final

This model is a fine-tuned version of facebook/dinov2-base on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.4955
• Accuracy: 0.9470
• F1 Weighted: 0.9478
• F1 Macro: 0.9478

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• 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
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Weighted	F1 Macro
0.306	1.0	44	0.5057	0.7819	0.7847	0.7847
0.1517	2.0	88	0.1859	0.9424	0.9422	0.9422
0.1301	3.0	132	0.2269	0.9315	0.9316	0.9316
0.0946	4.0	176	0.2158	0.9393	0.9402	0.9402
0.0846	5.0	220	0.2081	0.9455	0.9454	0.9454
0.0738	6.0	264	0.2057	0.9486	0.9491	0.9491
0.0388	7.0	308	0.3705	0.9128	0.9158	0.9158
0.025	8.0	352	0.3852	0.9221	0.9239	0.9239
0.0226	9.0	396	0.2083	0.9517	0.9522	0.9522
0.0126	10.0	440	0.3026	0.9408	0.9415	0.9415
0.0437	11.0	484	0.2995	0.9533	0.9538	0.9538
0.0193	12.0	528	0.4247	0.9377	0.9375	0.9375
0.0044	13.0	572	0.5568	0.9330	0.9351	0.9351
0.0015	14.0	616	0.4955	0.9470	0.9478	0.9478

Framework versions

• Transformers 4.47.0
• Pytorch 2.10.0+cu128
• Datasets 4.8.5
• Tokenizers 0.21.4