Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Feature/atari wrapper #92

Merged
merged 12 commits into from
Jan 3, 2022
Merged

Feature/atari wrapper #92

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
12 commits
Select commit Hold shift + click to select a range
966a276
implement wrapper
Dec 23, 2021
335cb15
update config and default value
ramanuzan Dec 23, 2021
3970e85
fix minor
Dec 23, 2021
a7c3d9c
Update conftest.py
ramanuzan Dec 23, 2021
a831f5b
update wrapper code
Dec 29, 2021
393ce9a
:art: Format Python code with psf/black (#95)
github-actions[bot] Dec 29, 2021
dc851ba
update skip frame buffer
Dec 29, 2021
d1f7f5a
update log file name %Y%m%d%H%M%S -> %Y%m%d%H%M%S%f
Jan 3, 2022
5d729a9
update denominator
Jan 3, 2022
1db9d56
update config
Jan 3, 2022
1b7e620
update fire reset assertion
Jan 3, 2022
a18c8fd
apply black
Jan 3, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions jorldy/config/ape_x/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/ape_x/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### Ape-X Pong_ML-Agents Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "ape_x",
Expand Down
3 changes: 2 additions & 1 deletion jorldy/config/ape_x/procgen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,8 @@
"render": False,
"gray_img": True,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
}

Expand Down
5 changes: 3 additions & 2 deletions jorldy/config/c51/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/c51/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### C51 Pong_ML-Agents Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "c51",
Expand Down
3 changes: 2 additions & 1 deletion jorldy/config/c51/procgen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,8 @@
"render": False,
"gray_img": True,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
}

Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/ddpg/hopper_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### DDPG Hopper Config ###

env = {"name": "hopper_mlagent", "train_mode": True}
env = {"name": "hopper_mlagent", "time_scale": 12.0}

agent = {
"name": "ddpg",
Expand Down
5 changes: 3 additions & 2 deletions jorldy/config/double/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/double/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### Double DQN Pong MLAgent Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "double",
Expand Down
3 changes: 2 additions & 1 deletion jorldy/config/double/procgen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,8 @@
"render": False,
"gray_img": True,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
}

Expand Down
13 changes: 7 additions & 6 deletions jorldy/config/dqn/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand All @@ -27,19 +28,19 @@
}

optim = {
"name": "rmsprop",
"lr": 2.5e-4,
"name": "adam",
"lr": 1e-4,
}

train = {
"training": True,
"load_path": None,
"run_step": 30000000,
"run_step": 10000000,
"print_period": 10000,
"save_period": 100000,
"eval_iteration": 5,
"record": True,
"record_period": 300000,
"record_period": 500000,
# distributed setting
"update_period": 32,
"num_workers": 16,
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/dqn/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### DQN Pong_ML-Agents Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "dqn",
Expand Down
3 changes: 2 additions & 1 deletion jorldy/config/dqn/procgen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,8 @@
"render": False,
"gray_img": True,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
}

Expand Down
5 changes: 3 additions & 2 deletions jorldy/config/dueling/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/dueling/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### Dueling DQN Pong_ML-Agents Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "dueling",
Expand Down
3 changes: 2 additions & 1 deletion jorldy/config/dueling/procgen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,8 @@
"render": False,
"gray_img": True,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
}

Expand Down
5 changes: 3 additions & 2 deletions jorldy/config/icm_ppo/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/icm_ppo/drone_delivery_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### ICM PPO Drone Delivery Config ###

env = {"name": "drone_delivery_mlagent", "train_mode": True}
env = {"name": "drone_delivery_mlagent", "time_scale": 12.0}

agent = {
"name": "icm_ppo",
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/icm_ppo/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### ICM PPO Pong_ML-Agents Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "icm_ppo",
Expand Down
3 changes: 2 additions & 1 deletion jorldy/config/icm_ppo/procgen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,8 @@
"render": False,
"gray_img": True,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
}

Expand Down
5 changes: 3 additions & 2 deletions jorldy/config/icm_ppo/super_mario_bros.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand Down
5 changes: 3 additions & 2 deletions jorldy/config/iqn/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/iqn/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### IQN Pong_ML-Agents Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "iqn",
Expand Down
3 changes: 2 additions & 1 deletion jorldy/config/iqn/procgen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,8 @@
"render": False,
"gray_img": True,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
}

Expand Down
5 changes: 3 additions & 2 deletions jorldy/config/m_dqn/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,8 +8,9 @@
"img_height": 84,
"stack_frame": 4,
"no_op": True,
"reward_clip": False,
"dead_penalty": False,
"skip_frame": 4,
"reward_clip": True,
"episodic_life": True,
}

agent = {
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/m_dqn/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### Munchausen DQN Pong_ML-Agents Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "m_dqn",
Expand Down
3 changes: 2 additions & 1 deletion jorldy/config/m_dqn/procgen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,8 @@
"render": False,
"gray_img": True,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
}

Expand Down
5 changes: 3 additions & 2 deletions jorldy/config/m_iqn/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/m_iqn/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### Munchausen IQN Pong_ML-Agents Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "m_iqn",
Expand Down
3 changes: 2 additions & 1 deletion jorldy/config/m_iqn/procgen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,8 @@
"render": False,
"gray_img": True,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
}

Expand Down
5 changes: 3 additions & 2 deletions jorldy/config/mpo/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/mpo/drone_delivery_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### MPO Drone Delivery Config ###

env = {"name": "drone_delivery_mlagent", "train_mode": True}
env = {"name": "drone_delivery_mlagent", "time_scale": 12.0}

agent = {
"name": "mpo",
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/mpo/hopper_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### MPO Hopper Config ###

env = {"name": "hopper_mlagent", "train_mode": True}
env = {"name": "hopper_mlagent", "time_scale": 12.0}

agent = {
"name": "mpo",
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/mpo/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### MPO Pong_ML-Agents Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "mpo",
Expand Down
3 changes: 2 additions & 1 deletion jorldy/config/mpo/procgen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,8 @@
"render": False,
"gray_img": True,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
}

Expand Down
5 changes: 3 additions & 2 deletions jorldy/config/multistep/atari.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,10 @@
"img_width": 84,
"img_height": 84,
"stack_frame": 4,
"no_op": False,
"no_op": True,
"skip_frame": 4,
"reward_clip": True,
"dead_penalty": True,
"episodic_life": True,
}

agent = {
Expand Down
2 changes: 1 addition & 1 deletion jorldy/config/multistep/pong_mlagent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
### Multistep DQN Pong_ML-Agents Config ###

env = {"name": "pong_mlagent", "train_mode": True}
env = {"name": "pong_mlagent", "time_scale": 12.0}

agent = {
"name": "multistep",
Expand Down
Loading