README.md

metadata

license: other
base_model: nvidia/segformer-b3-finetuned-ade-512-512
tags:
  - generated_from_keras_callback
model-index:
  - name: chribark/segformer-b3-finetuned-ade-512-512-finetuned-UAVid
    results: []

chribark/segformer-b3-finetuned-ade-512-512-finetuned-UAVid

This model is a fine-tuned version of nvidia/segformer-b3-finetuned-ade-512-512 on an unknown dataset. It achieves the following results on the evaluation set:

• Train Loss: 0.3207
• Validation Loss: 0.3852
• Validation Mean Iou: 0.0093
• Validation Mean Accuracy: 0.0378
• Validation Overall Accuracy: 0.0336
• Validation Accuracy Clutter: 0.0293
• Validation Accuracy Building: 0.0228
• Validation Accuracy Road: 0.0896
• Validation Accuracy Static Car: 0.0006
• Validation Accuracy Tree: 0.1118
• Validation Accuracy Vegetation: 0.0103
• Validation Accuracy Human: 0.0
• Validation Accuracy Moving Car: nan
• Validation Iou Clutter: 0.0199
• Validation Iou Building: 0.0037
• Validation Iou Road: 0.0072
• Validation Iou Static Car: 0.0006
• Validation Iou Tree: 0.0427
• Validation Iou Vegetation: 0.0000
• Validation Iou Human: 0.0
• Validation Iou Moving Car: 0.0
• Epoch: 9

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', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': 6e-05, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Validation Loss	Validation Mean Iou	Validation Mean Accuracy	Validation Overall Accuracy	Validation Accuracy Clutter	Validation Accuracy Building	Validation Accuracy Road	Validation Accuracy Static Car	Validation Accuracy Tree	Validation Accuracy Vegetation	Validation Accuracy Human	Validation Accuracy Moving Car	Validation Iou Clutter	Validation Iou Building	Validation Iou Road	Validation Iou Static Car	Validation Iou Tree	Validation Iou Vegetation	Validation Iou Human	Validation Iou Moving Car	Epoch
1.2150	0.6691	0.0149	0.0576	0.0526	0.0355	0.0698	0.1084	0.0000	0.1892	0.0	0.0	nan	0.0241	0.0108	0.0100	0.0000	0.0739	0.0	0.0	0.0	0
0.6829	0.5768	0.0213	0.0817	0.0809	0.0750	0.0083	0.2008	0.0	0.2878	0.0	0.0	nan	0.0501	0.0014	0.0151	0.0	0.1035	0.0	0.0	0.0	1
0.5679	0.4969	0.0151	0.0710	0.0516	0.0295	0.0240	0.2164	0.0000	0.2230	0.0041	0.0	nan	0.0206	0.0039	0.0151	0.0000	0.0814	0.0000	0.0	0.0	2
0.5010	0.4654	0.0135	0.0553	0.0499	0.0437	0.0160	0.1503	0.0000	0.1773	0.0	0.0	nan	0.0301	0.0026	0.0096	0.0000	0.0660	0.0	0.0	0.0	3
0.4501	0.4507	0.0107	0.0504	0.0370	0.0343	0.0176	0.1737	0.0001	0.1191	0.0082	0.0	nan	0.0228	0.0028	0.0130	0.0001	0.0469	0.0000	0.0	0.0	4
0.4229	0.4257	0.0116	0.0436	0.0445	0.0515	0.0120	0.1162	0.0002	0.1222	0.0031	0.0	nan	0.0344	0.0020	0.0087	0.0002	0.0471	0.0000	0.0	0.0	5
0.3823	0.4131	0.0127	0.0504	0.0455	0.0374	0.0124	0.1251	0.0004	0.1705	0.0072	0.0	nan	0.0251	0.0020	0.0098	0.0004	0.0639	0.0000	0.0	0.0	6
0.3610	0.4006	0.0121	0.0518	0.0421	0.0300	0.0162	0.1272	0.0004	0.1675	0.0215	0.0	nan	0.0207	0.0026	0.0096	0.0004	0.0631	0.0001	0.0	0.0	7
0.3428	0.3923	0.0119	0.0465	0.0430	0.0327	0.0166	0.0977	0.0003	0.1670	0.0113	0.0	nan	0.0223	0.0027	0.0078	0.0003	0.0617	0.0000	0.0	0.0	8
0.3207	0.3852	0.0093	0.0378	0.0336	0.0293	0.0228	0.0896	0.0006	0.1118	0.0103	0.0	nan	0.0199	0.0037	0.0072	0.0006	0.0427	0.0000	0.0	0.0	9

Framework versions

• Transformers 4.40.2
• TensorFlow 2.15.0
• Datasets 2.19.1
• Tokenizers 0.19.1