Update readme.md

Author: zylo117

readme.md

@@ -149,15 +149,15 @@ Check out this [tutorial](tutorial/train_shape.ipynb) if you are new to this. Yo
 ### 3.c. Train a custom dataset with pretrained weights (Highly Recommended)
 
     # train efficientdet-d2 on a custom dataset with pretrained weights
-    # with batchsize 8 and learning rate 1e-5 for 10 epoches
+    # with batchsize 8 and learning rate 1e-3 for 10 epoches
 
-    python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-5 --num_epochs 10 \
+    python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-3 --num_epochs 10 \
      --load_weights /path/to/your/weights/efficientdet-d2.pth
 
     # with a coco-pretrained, you can even freeze the backbone and train heads only
     # to speed up training and help convergence.
 
-    python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-5 --num_epochs 10 \
+    python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-3 --num_epochs 10 \
      --load_weights /path/to/your/weights/efficientdet-d2.pth \
      --head_only True
 
@@ -170,7 +170,7 @@ Check out this [tutorial](tutorial/train_shape.ipynb) if you are new to this. Yo
 
     # let says you started a training session like this.
 
-    python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-5 \
+    python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-3 \
      --load_weights /path/to/your/weights/efficientdet-d2.pth \
      --head_only True
      
@@ -179,7 +179,7 @@ Check out this [tutorial](tutorial/train_shape.ipynb) if you are new to this. Yo
     # now you want to resume training from the last checkpoint
     # simply set load_weights to 'last'
 
-    python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-5 \
+    python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-3 \
      --load_weights last \
      --head_only True
 
@@ -193,7 +193,7 @@ Check out this [tutorial](tutorial/train_shape.ipynb) if you are new to this. Yo
 ### 7. Debug training (optional)
 
     # when you get bad result, you need to debug the training result.
-    python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-5 --debug True
+    python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-3 --debug True
 
     # then checkout test/ folder, there you can visualize the predicted boxes during training
     # don't panic if you see countless of error boxes, it happens when the training is at early stage.