Image Segmentation
TerraTorch
English
Pytorch
mmsegmentation
segmentation
burn scars
Geospatial
Foundation model
Instructions to use ibm-nasa-geospatial/Prithvi-EO-1.0-100M-burn-scar with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- TerraTorch
How to use ibm-nasa-geospatial/Prithvi-EO-1.0-100M-burn-scar with TerraTorch:
from terratorch.registry import BACKBONE_REGISTRY model = BACKBONE_REGISTRY.build("ibm-nasa-geospatial/Prithvi-EO-1.0-100M-burn-scar") - Notebooks
- Google Colab
- Kaggle
Commit ·
fed16ce
1
Parent(s): 67b68ed
fix bug
Browse files
burn_scars_Prithvi_100M.py
CHANGED
|
@@ -218,6 +218,6 @@ model = dict(
|
|
| 218 |
align_corners=False,
|
| 219 |
loss_decode=loss_func),
|
| 220 |
train_cfg=dict(),
|
| 221 |
-
test_cfg=dict(mode='slide', stride=(tile_size/2, tile_size/2), crop_size=(tile_size, tile_size)))
|
| 222 |
gpu_ids = range(0, 1)
|
| 223 |
auto_resume = False
|
|
|
|
| 218 |
align_corners=False,
|
| 219 |
loss_decode=loss_func),
|
| 220 |
train_cfg=dict(),
|
| 221 |
+
test_cfg=dict(mode='slide', stride=(int(tile_size/2), int(tile_size/2)), crop_size=(tile_size, tile_size)))
|
| 222 |
gpu_ids = range(0, 1)
|
| 223 |
auto_resume = False
|