AGarioud commited on
Commit
a0342ab
1 Parent(s): 0f22040

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +20 -19
README.md CHANGED
@@ -108,7 +108,7 @@ pipeline_tag: image-segmentation
108
  - **Compute infrastructure:**
109
  - software: python, pytorch-lightning
110
  - hardware: HPC/AI resources provided by GENCI-IDRIS
111
- - **License:** : Apache 2.0
112
 
113
  ---
114
 
@@ -190,24 +190,25 @@ Statistics of the TRAIN+VALIDATION set :
190
 
191
  #### Training Hyperparameters
192
 
193
- * Model architecture: FPN (implementation from the [Segmentation Models Pytorch library](https://segmentation-modelspytorch.readthedocs.io/en/latest/docs/api.html#unet))
194
- * Encoder : Resnet-34 pre-trained with ImageNet
195
- * Augmentation :
196
- * VerticalFlip(p=0.5)
197
- * HorizontalFlip(p=0.5)
198
- * RandomRotate90(p=0.5)
199
- * Input normalization (mean=0 | std=1):
200
- * norm_means: [105.08, 110.87, 101.82, 106.38, 53.26]
201
- * norm_stds: [52.17, 45.38, 44, 39.69, 79.3]
202
- * Seed: 2022
203
- * Batch size: 10
204
- * Number of epochs : 200
205
- * Early stopping : patience 30 and val_loss as monitor criterium
206
- * Optimizer : SGD
207
- * Schaeduler : mode = "min", factor = 0.5, patience = 10, cooldown = 4, min_lr = 1e-7
208
- * Learning rate : 0.02
209
- * Class Weights : [1-building: 1.0 , 2-pervious surface: 1.0 , 3-impervious surface: 1.0 , 4-bare soil: 1.0 , 5-water: 1.0 , 6-coniferous: 1.0 , 7-deciduous: 1.0 , 8-brushwood: 1.0 , 9-vineyard: 1.0 , 10-herbaceous vegetation: 1.0 , 11-agricultural land: 1.0 , 12-plowed land: 1.0 , 13-swimming_pool: 1.0 , 14-snow: 1.0 , 15-clear cut: 0.0 , 16-mixed: 0.0 , 17-ligneous: 0.0 , 18-greenhouse: 1.0 , 19-other: 0.0]
210
-
 
211
 
212
  #### Speeds, Sizes, Times
213
 
 
108
  - **Compute infrastructure:**
109
  - software: python, pytorch-lightning
110
  - hardware: HPC/AI resources provided by GENCI-IDRIS
111
+ - **License:** : Etalab 2.0
112
 
113
  ---
114
 
 
190
 
191
  #### Training Hyperparameters
192
 
193
+ ```yaml
194
+ - Model architecture: FPN #(implementation from the [Segmentation Models Pytorch library](https://segmentation-modelspytorch.readthedocs.io/en/latest/docs/api.html#unet))
195
+ - Encoder : Resnet-34 pre-trained with ImageNet
196
+ - Augmentation :
197
+ - VerticalFlip(p=0.5)
198
+ - HorizontalFlip(p=0.5)
199
+ - RandomRotate90(p=0.5)
200
+ - Input normalization (mean=0 | std=1):
201
+ - norm_means: [105.08, 110.87, 101.82, 106.38, 53.26]
202
+ - norm_stds: [52.17, 45.38, 44, 39.69, 79.3]
203
+ - Seed: 2022
204
+ - Batch size: 10
205
+ - Number of epochs : 200
206
+ - Early stopping : patience 30 and val_loss as monitor criterium
207
+ - Optimizer : SGD
208
+ - Schaeduler : mode = "min", factor = 0.5, patience = 10, cooldown = 4, min_lr = 1e-7
209
+ - Learning rate : 0.02
210
+ - Class Weights : [1-building: 1.0 , 2-pervious surface: 1.0 , 3-impervious surface: 1.0 , 4-bare soil: 1.0 , 5-water: 1.0 , 6-coniferous: 1.0 , 7-deciduous: 1.0 , 8-brushwood: 1.0 , 9-vineyard: 1.0 , 10-herbaceous vegetation: 1.0 , 11-agricultural land: 1.0 , 12-plowed land: 1.0 , 13-swimming_pool: 1.0 , 14-snow: 1.0 , 15-clear cut: 0.0 , 16-mixed: 0.0 , 17-ligneous: 0.0 , 18-greenhouse: 1.0 , 19-other: 0.0]
211
+ ```
212
 
213
  #### Speeds, Sizes, Times
214