Skip to content
GitLab
Projects
Groups
Snippets
Help
Loading...
Help
What's new
7
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
Open sidebar
sjmonagi
Simulated visual based reinforcement learning for navigation with Hindsight experience Replay
Commits
0bd8fcab
Commit
0bd8fcab
authored
Apr 10, 2020
by
sjmonagi
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
Update __Manual to run
parent
37f1c80d
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
24 additions
and
7 deletions
+24
-7
__Manual to run
__Manual to run
+24
-7
No files found.
__Manual to run
View file @
0bd8fcab
...
...
@@ -8,23 +8,40 @@
### Steps to run the code
1- load the Autoencoder weights
class Network(object):
#### Create model
def __init__(self):
self.save_path = '### YOUR PATH to AUTOENCODER WEIGHTS###/CNN_AE.ckpt'
file name: autoencoder
class Network(object):
#### Create model
21- def __init__(self):
22- self.save_path = '### YOUR PATH to AUTOENCODER WEIGHTS###/CNN_AE.ckpt'
3- setting the directory where the training weights to be save -> dir = /your_directory/
2- The model was trained on Environment "FloorPlan225" which is the default in the Environment Class
### Any changes to the Environment.py specially Scene and Actions, those changes need to be done also in Environment_top_view.py file
Running file in any folder those settings need to be modified
running files path:
for playing the trajectory:
playing/trajectory_plotting_images_only.py
playing/trajectory_plotting_images_pos.py
playing/trajectory_plotting_shaped_reward_only.py
3- Specify which reward type to use in DRQN_HER_Training.py -> reward = "sparse" # reward type "shaped","sparse"
for training and evalation shaped reward only:
Agent_shaped_reward_only/DRQN_HER_Training.py
4- For Random Agent initialization from random position only -> random_init_position = True
for training and evalation images and position images only:
Agent_image_only/DRQN_HER_Training.py
for training and evalation images and positions images and positions:
Agent_shaped_reward_wiz_her_images_and_postions/DRQN_HER_Training.py
3- Specify which reward type to use in DRQN_HER_Training.py -> reward = ["shaped"],["sparse"]
lines 25 For Random Agent initialization from random position only -> random_init_position = True
random_init_pose = False
5- For
Random Agent initialization from random position and Pose -> random_init_position = False
lines 22
Random Agent initialization from random position and Pose -> random_init_position = False
random_init_pose = True
6- Top enable or disable the top view render -> top_view = True/False # displaying top-view
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment