Downloads · 30 days
0
segformer-b0-kaggle-ktm-dataset-block-fold2
Pranilllllll/segformer-b0-kaggle-ktm-dataset-block-fold2
segformer-b0-kaggle-ktm-dataset-block-fold2 is a machine learning model from Pranilllllll. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers.
Kathmandu Valley land-use segmentation, 6 classes, ignoreindex=255.
Access
Public
Updated Aug 18, 2026
Repo size
2.7 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt74.7 MB · 95%
At a glance
- Library
- transformers
- Access
- Public
- Created
- Aug 18, 2026
- Updated
- Aug 18, 2026
- SHA
- b85dd12a
Model card
From the Hugging Face model README
GeoNUSAF - SegFormer-B0 - block split, fold 2
Kathmandu Valley land-use segmentation, 6 classes, ignore_index=255.
| field | value |
|---|---|
| architecture | SegformerForSemanticSegmentation, nvidia/segformer-b0-finetuned-ade-512-512 |
| split mode | block |
| fold | 2 of 3 |
| seed | 42 |
| input | 512x512, ImageNet norm, effective GSD 0.586 m/px |
| classes | Residential, Road, River, Forest, UnusedLand, Agricultural |
| lr (head/encoder) | 6e-05 / 6e-06 |
| regularization | wd 0.01, drop_path 0.1, smooth 0.05, EMA True |
| best epoch | 138 |
| val mIoU | 0.4351 |
| val mF1 | 0.5699 |
| val OA | 0.8315 |
| val kappa | 0.6232 |
Per-class (validation)
| class | IoU | F1 |
|---|---|---|
| Residential | 0.8665 | 0.9285 |
| Road | 0.3214 | 0.4865 |
| River | 0.1395 | 0.2449 |
| Forest | 0.4771 | 0.6460 |
| UnusedLand | 0.2478 | 0.3972 |
| Agricultural | 0.5583 | 0.7165 |
Checkpoint best.pt holds model_state (EMA weights when EMA is on), the run cfg, the full HF model_config, and metrics.
- Library
- transformers
- Created
- Aug 18, 2026
- Updated
- Aug 18, 2026