Update readme.md

docs/TESTS/Test2_DeepLabV3+/readme.md

@@ -8,17 +8,22 @@
 
 #### 1. Pretrained model:
       pretrained_models/model_13_2_2_2_epoch_580.pth: Trained for 580 epochs on Cityscapes train and 3333 + 745 images from Berkeley DeepDrive.
+---
 #### 2. Dataset preprocesing:
       Run: 1_preprocess_data.ipynb   (ONLY NEED TO DO THIS ONCE!)
+---
 #### 3. Train model on Cityscapes:
       Run: 2_train.ipynb
+---
 #### 4. Evaluation
 ##### eval_on_val
       Run: 3_eval_on_val.ipynb
 This will run the pretrained model on all images in Cityscapes val, compute and print the loss, and save the predicted segmentation images in deeplabv3/training_logs/model_eval_val.
+
 ##### eval_on_val_for_metrics
       Run: 4_eval_on_val_for_metrics.ipynb
 This will run the pretrained model on all images in Cityscapes val, upsample the predicted segmentation images to the original Cityscapes image size (1024, 2048), and compute and print performance metrics.
+---
 #### 5. Visualization
 ##### run_on_seq
       Run: 5_run_on_seq.ipynb