unet-blood-vessels

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

unet-blood-vessels

This model is a fine-tuned version of on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5415
• Dice: 0.6738
• Iou: 0.7312
• Iou Fg: 0.5355
• Iou Bg: 0.9269
• Acc: 0.9326
• Recall: 0.8356
• F1: 0.6975

Model description

Simple Unet with an efficientnet backbone and double convolutional layers with GELU activation.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0003
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-10 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 100.0

Training results

Metric Plots

Best Metrics with TTA

Best threshold: 0.600 Metrics @ best: {'dice': 0.7461113333702087, 'iou': 0.7712731957435608, 'iou_fg': 0.5950379967689514, 'iou_bg': 0.9475083947181702}

Inference Results

Framework versions

• Transformers 4.57.6
• Pytorch 2.8.0+cu128
• Datasets 4.5.0
• Tokenizers 0.22.2