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

Safe gpr params #130

Closed
wants to merge 12 commits into from
Closed

Safe gpr params #130

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
e5b3764
test larger kps
budzianowski Dec 5, 2024
adb0ead
new kps stable walk
budzianowski Dec 6, 2024
9b5e1b7
save trained policy
budzianowski Dec 6, 2024
0d5700d
add force and trimesh
budzianowski Dec 6, 2024
4a01fc4
fix default standing
WT-MM Dec 6, 2024
0dffaa9
Merge branch 'master' of github.com:kscalelabs/sim into refine_gpr
budzianowski Dec 9, 2024
6e113ce
switch to robot
WT-MM Dec 9, 2024
57fcfea
update values
budzianowski Dec 9, 2024
73dee0c
update to even lower efforts
budzianowski Dec 9, 2024
2cf4840
new params
WT-MM Dec 9, 2024
fd57e2b
0.5s cycle
WT-MM Dec 9, 2024
bc4b9d3
increase kps
WT-MM Dec 9, 2024
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
Binary file modified examples/walking_policy.pt
View file
Open in desktop
Binary file not shown.
12 changes: 6 additions & 6 deletions sim/envs/humanoids/gpr_config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,7 +30,7 @@ class safety(LeggedRobotCfg.safety):
# safety factors
pos_limit = 1.0
vel_limit = 1.0
torque_limit = 0.85
torque_limit = 1.0

class asset(LeggedRobotCfg.asset):
name = "gpr"
Expand Down Expand Up @@ -123,7 +123,7 @@ class domain_rand(LeggedRobotCfg.domain_rand):

randomize_base_mass = True
added_mass_range = [-3.0, 3.0]
push_robots = False
push_robots = True
push_interval_s = 4
max_push_vel_xy = 0.2
max_push_ang_vel = 0.4
Expand All @@ -150,14 +150,14 @@ class rewards:
max_dist = 0.5
# put some settings here for LLM parameter tuning
target_joint_pos_scale = 0.17 # rad
target_feet_height = 0.06 # m
target_feet_height = 0.05 # m

cycle_time = 0.64 # sec
cycle_time = 0.5 # sec
# if true negative total rewards are clipped at zero (avoids early termination problems)
only_positive_rewards = True
# tracking reward = exp(error*sigma)
tracking_sigma = 5.0
max_contact_force = 700 # forces above this value are penalized
max_contact_force = 400 # forces above this value are penalized

class scales:
# reference motion tracking
Expand Down Expand Up @@ -213,7 +213,7 @@ class GprStandingCfg(GprCfg):

class rewards:
# quite important to keep it right
base_height_target = 1.06
base_height_target = Robot.height
min_dist = 0.2
max_dist = 0.5
# put some settings here for LLM parameter tuning
Expand Down
3 changes: 3 additions & 0 deletions sim/model_export.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -242,6 +242,8 @@ def get_actor_policy(model_path: str, cfg: ActorCfg) -> Tuple[nn.Module, dict, T
num_actions = a_model.num_actions
num_observations = a_model.num_observations

default_standing = list(a_model.robot.default_standing().values())

return (
a_model,
{
Expand All @@ -250,6 +252,7 @@ def get_actor_policy(model_path: str, cfg: ActorCfg) -> Tuple[nn.Module, dict, T
"robot_damping": robot_damping,
"num_actions": num_actions,
"num_observations": num_observations,
"default_standing": default_standing,
},
input_tensors,
)
Expand Down
30 changes: 15 additions & 15 deletions sim/resources/gpr/joints.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -156,33 +156,33 @@ def default_limits(cls) -> Dict[str, Dict[str, float]]:
@classmethod
def stiffness(cls) -> Dict[str, float]:
return {
"hip_y": 120,
"hip_x": 60,
"hip_z": 60,
"knee": 120,
"ankle_y": 17,
"hip_y": 100, # 04
"hip_x": 40, # 03
"hip_z": 40, # 03
"knee": 100, # 04
"ankle_y": 17, # 02
}

# d_gains
@classmethod
def damping(cls) -> Dict[str, float]:
return {
"hip_y": 10,
"hip_x": 5,
"hip_z": 5,
"knee": 10,
"ankle_y": 5,
"hip_y": 10, # 04
"hip_x": 10, # 03
"hip_z": 10, # 03
"knee": 10, # 04
"ankle_y": 5, # 02
}

# # pos_limits
@classmethod
def effort(cls) -> Dict[str, float]:
return {
"hip_y": 120,
"hip_x": 60,
"hip_z": 60,
"knee": 120,
"ankle_y": 17,
"hip_y": 30, # 04
"hip_x": 20, # 03
"hip_z": 20, # 03
"knee": 30, # 04
"ankle_y": 17, # 02
}

# # vel_limits
Expand Down
18 changes: 9 additions & 9 deletions sim/resources/gpr/robot_fixed.urdf
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -95,7 +95,7 @@
<origin xyz="-0.008010608503738428 -0.43201043179234816 0.0877000068648673" rpy="-3.1415925071795874 4.641020678874952e-08 -3.1415926535897913" />
<parent link="body1-part" />
<child link="leg0_shell" />
<limit effort="120" velocity="10" lower="-1.5707963" upper="1.5707963" />
<limit effort="30" velocity="10" lower="-1.5707963" upper="1.5707963" />
<axis xyz="0 0 -1" />
</joint>
<link name="leg0_shell">
Expand Down Expand Up @@ -124,7 +124,7 @@
<origin xyz="-0.008010600340183055 -0.4320104467591277 -0.08819999313513187" rpy="9.99999991702083e-08 4.641020678874952e-08 -3.1415926535897913" />
<parent link="body1-part" />
<child link="leg0_shell_2" />
<limit effort="120" velocity="10" lower="-1.5707963" upper="1.5707963" />
<limit effort="30" velocity="10" lower="-1.5707963" upper="1.5707963" />
<axis xyz="0 0 -1" />
</joint>
<link name="leg0_shell_2">
Expand Down Expand Up @@ -211,7 +211,7 @@
<origin xyz="0.02649999999999997 -1.5165911463888016e-09 -0.06950000151659115" rpy="-1.5707963 0.0 1.5707963" />
<parent link="leg0_shell" />
<child link="leg1_shell" />
<limit effort="60" velocity="10" lower="-0.34906585" upper="3.1852259" />
<limit effort="20" velocity="10" lower="-0.34906585" upper="3.1852259" />
<axis xyz="0 0 1" />
</joint>
<link name="leg1_shell">
Expand Down Expand Up @@ -240,7 +240,7 @@
<origin xyz="0.02649999999999997 -1.5165911463888016e-09 -0.06950000151659115" rpy="1.5707963 -4.641020634466031e-08 -1.5707963535897922" />
<parent link="leg0_shell_2" />
<child link="leg1_shell3" />
<limit effort="60" velocity="10" lower="-0.34906585" upper="3.1852259" />
<limit effort="20" velocity="10" lower="-0.34906585" upper="3.1852259" />
<axis xyz="0 0 1" />
</joint>
<link name="leg1_shell3">
Expand Down Expand Up @@ -327,7 +327,7 @@
<origin xyz="-0.15649999999999997 0.0 0.027499998483408852" rpy="-0.0 1.5707963 0.0" />
<parent link="leg1_shell" />
<child link="leg2_shell" />
<limit effort="60" velocity="10" lower="-1.5707963" upper="1.5707963" />
<limit effort="20" velocity="10" lower="-1.5707963" upper="1.5707963" />
<axis xyz="0 0 1" />
</joint>
<link name="leg2_shell">
Expand Down Expand Up @@ -356,7 +356,7 @@
<origin xyz="-0.15649999999999997 0.0 0.027499998483408852" rpy="-3.1415926535897922 -1.5707962732050302 0.0" />
<parent link="leg1_shell3" />
<child link="leg2_shell_2" />
<limit effort="60" velocity="10" lower="-1.5707963" upper="1.5707963" />
<limit effort="20" velocity="10" lower="-1.5707963" upper="1.5707963" />
<axis xyz="0 0 1" />
</joint>
<link name="leg2_shell_2">
Expand Down Expand Up @@ -443,7 +443,7 @@
<origin xyz="0.0 0.0342 -0.14250000009378214" rpy="1.5707963000000003 0.0 0.0" />
<parent link="leg2_shell" />
<child link="leg3_shell2" />
<limit effort="120" velocity="10" lower="-1.57" upper="0" />
<limit effort="30" velocity="10" lower="-1.57" upper="0" />
<axis xyz="0 0 1" />
</joint>
<link name="leg3_shell2">
Expand Down Expand Up @@ -472,7 +472,7 @@
<origin xyz="0.0 0.0342 -0.14250000009378214" rpy="1.5707963000000003 0.0 0.0" />
<parent link="leg2_shell_2" />
<child link="leg3_shell22" />
<limit effort="120" velocity="10" lower="0" upper="1.57" />
<limit effort="30" velocity="10" lower="0" upper="1.57" />
<axis xyz="0 0 1" />
</joint>
<link name="leg3_shell22">
Expand Down Expand Up @@ -613,4 +613,4 @@
<origin xyz="0.021026404502789243 -0.018472020400304597 -0.01930430937248649" rpy="0 0 0" />
</inertial>
</link>
</robot>
</robot>
4 changes: 2 additions & 2 deletions sim/sim2sim.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -319,7 +319,7 @@ def run_mujoco(

policy_cfg = ActorCfg(embodiment=args.embodiment)
if args.embodiment == "gpr":
policy_cfg.cycle_time = 0.25
policy_cfg.cycle_time = 0.5
cfg = Sim2simCfg(
sim_duration=10.0,
dt=0.001,
Expand All @@ -337,7 +337,7 @@ def run_mujoco(
cycle_time=policy_cfg.cycle_time,
)

if args.load_model.endswith(".kinfer"):
if args.load_model.endswith(".kinfer") or args.load_model.endswith(".onnx"):
policy = ONNXModel(args.load_model)
else:
actor_model, sim2sim_info, input_tensors = get_actor_policy(args.load_model, policy_cfg)
Expand Down
Loading