README.md

metadata

license: other
tags:
  - generated_from_keras_callback
model-index:
  - name: AhamadShaik/SegFormer_RESIZE_NLM
    results: []

AhamadShaik/SegFormer_RESIZE_NLM

This model is a fine-tuned version of nvidia/mit-b0 on an unknown dataset. It achieves the following results on the evaluation set:

• Train Loss: 0.0886
• Train Dice Coef: 0.7856
• Train Iou: 0.6535
• Validation Loss: 0.0584
• Validation Dice Coef: 0.8603
• Validation Iou: 0.7566
• Train Lr: 1e-04
• Epoch: 5

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:

• optimizer: {'name': 'Adam', 'learning_rate': 1e-04, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Train Dice Coef	Train Iou	Validation Loss	Validation Dice Coef	Validation Iou	Train Lr	Epoch
0.2282	0.5657	0.4102	0.1322	0.6524	0.4967	1e-04	0
0.1354	0.6853	0.5329	0.0855	0.7853	0.6544	1e-04	1
0.1105	0.7364	0.5924	0.0737	0.8147	0.6916	1e-04	2
0.0985	0.7610	0.6226	0.0632	0.8518	0.7440	1e-04	3
0.0933	0.7745	0.6399	0.0627	0.8455	0.7351	1e-04	4
0.0886	0.7856	0.6535	0.0584	0.8603	0.7566	1e-04	5

Framework versions

• Transformers 4.27.4
• TensorFlow 2.10.1
• Datasets 2.11.0
• Tokenizers 0.13.3