v1.0.0; max train step 120k

This commit is contained in:
wty-yy
2026-01-13 12:37:05 +08:00
parent b4bf32b4cd
commit 02da83d800
6 changed files with 32 additions and 29 deletions

View File

@@ -1,3 +1,6 @@
# 20260113
## v1.0.0
1. 确定最大训练步数为120k
# 20260112 # 20260112
## v0.1.9 ## v0.1.9
1. 添加训练完成时, 等待RoboGauge全部评估完全后结束 1. 添加训练完成时, 等待RoboGauge全部评估完全后结束

View File

@@ -238,7 +238,7 @@ class GO2CfgPPO(LeggedRobotCfgPPO):
class runner(LeggedRobotCfgPPO.runner): class runner(LeggedRobotCfgPPO.runner):
run_name = '' run_name = ''
experiment_name = 'go2_ppo' experiment_name = 'go2_ppo'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgCTS(LeggedRobotCfgCTS): class GO2CfgCTS(LeggedRobotCfgCTS):
@@ -246,7 +246,7 @@ class GO2CfgCTS(LeggedRobotCfgCTS):
num_steps_per_env = 24 num_steps_per_env = 24
run_name = '' run_name = ''
experiment_name = 'go2_cts' experiment_name = 'go2_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class policy(LeggedRobotCfgCTS.policy): class policy(LeggedRobotCfgCTS.policy):
@@ -264,7 +264,7 @@ class GO2CfgMoECTS(LeggedRobotCfgMoECTS):
class runner(LeggedRobotCfgMoECTS.runner): class runner(LeggedRobotCfgMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_moe_cts' experiment_name = 'go2_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS): class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS):
@@ -275,7 +275,7 @@ class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS):
class runner(LeggedRobotCfgMCPCTS.runner): class runner(LeggedRobotCfgMCPCTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_mcp_cts' experiment_name = 'go2_mcp_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS): class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS):
@@ -285,7 +285,7 @@ class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS):
class runner(LeggedRobotCfgACMoECTS.runner): class runner(LeggedRobotCfgACMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_ac_moe_cts' experiment_name = 'go2_ac_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS): class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS):
@@ -295,7 +295,7 @@ class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS):
class runner(LeggedRobotCfgDualMoECTS.runner): class runner(LeggedRobotCfgDualMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_dual_moe_cts' experiment_name = 'go2_dual_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgREMCTS(LeggedRobotCfgREMCTS): class GO2CfgREMCTS(LeggedRobotCfgREMCTS):
@@ -305,5 +305,5 @@ class GO2CfgREMCTS(LeggedRobotCfgREMCTS):
class runner(LeggedRobotCfgREMCTS.runner): class runner(LeggedRobotCfgREMCTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_rem_cts' experiment_name = 'go2_rem_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500

View File

@@ -268,7 +268,7 @@ class GO2CfgPPO(LeggedRobotCfgPPO):
class runner(LeggedRobotCfgPPO.runner): class runner(LeggedRobotCfgPPO.runner):
run_name = '' run_name = ''
experiment_name = 'go2_ppo' experiment_name = 'go2_ppo'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgCTS(LeggedRobotCfgCTS): class GO2CfgCTS(LeggedRobotCfgCTS):
@@ -276,7 +276,7 @@ class GO2CfgCTS(LeggedRobotCfgCTS):
num_steps_per_env = 24 num_steps_per_env = 24
run_name = '' run_name = ''
experiment_name = 'go2_cts' experiment_name = 'go2_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class policy(LeggedRobotCfgCTS.policy): class policy(LeggedRobotCfgCTS.policy):
@@ -294,7 +294,7 @@ class GO2CfgMoECTS(LeggedRobotCfgMoECTS):
class runner(LeggedRobotCfgMoECTS.runner): class runner(LeggedRobotCfgMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_moe_cts' experiment_name = 'go2_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS): class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS):
@@ -305,7 +305,7 @@ class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS):
class runner(LeggedRobotCfgMCPCTS.runner): class runner(LeggedRobotCfgMCPCTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_mcp_cts' experiment_name = 'go2_mcp_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS): class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS):
@@ -315,7 +315,7 @@ class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS):
class runner(LeggedRobotCfgACMoECTS.runner): class runner(LeggedRobotCfgACMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_ac_moe_cts' experiment_name = 'go2_ac_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS): class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS):
@@ -325,7 +325,7 @@ class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS):
class runner(LeggedRobotCfgDualMoECTS.runner): class runner(LeggedRobotCfgDualMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_dual_moe_cts' experiment_name = 'go2_dual_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgREMCTS(LeggedRobotCfgREMCTS): class GO2CfgREMCTS(LeggedRobotCfgREMCTS):
@@ -335,5 +335,5 @@ class GO2CfgREMCTS(LeggedRobotCfgREMCTS):
class runner(LeggedRobotCfgREMCTS.runner): class runner(LeggedRobotCfgREMCTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_rem_cts' experiment_name = 'go2_rem_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500

View File

@@ -253,7 +253,7 @@ class GO2CfgPPO(LeggedRobotCfgPPO):
class runner(LeggedRobotCfgPPO.runner): class runner(LeggedRobotCfgPPO.runner):
run_name = '' run_name = ''
experiment_name = 'go2_ppo' experiment_name = 'go2_ppo'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgCTS(LeggedRobotCfgCTS): class GO2CfgCTS(LeggedRobotCfgCTS):
@@ -261,7 +261,7 @@ class GO2CfgCTS(LeggedRobotCfgCTS):
num_steps_per_env = 24 num_steps_per_env = 24
run_name = '' run_name = ''
experiment_name = 'go2_cts' experiment_name = 'go2_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class policy(LeggedRobotCfgCTS.policy): class policy(LeggedRobotCfgCTS.policy):
@@ -279,7 +279,7 @@ class GO2CfgMoECTS(LeggedRobotCfgMoECTS):
class runner(LeggedRobotCfgMoECTS.runner): class runner(LeggedRobotCfgMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_moe_cts' experiment_name = 'go2_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS): class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS):
@@ -290,7 +290,7 @@ class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS):
class runner(LeggedRobotCfgMCPCTS.runner): class runner(LeggedRobotCfgMCPCTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_mcp_cts' experiment_name = 'go2_mcp_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS): class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS):
@@ -300,7 +300,7 @@ class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS):
class runner(LeggedRobotCfgACMoECTS.runner): class runner(LeggedRobotCfgACMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_ac_moe_cts' experiment_name = 'go2_ac_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS): class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS):
@@ -310,7 +310,7 @@ class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS):
class runner(LeggedRobotCfgDualMoECTS.runner): class runner(LeggedRobotCfgDualMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_dual_moe_cts' experiment_name = 'go2_dual_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgREMCTS(LeggedRobotCfgREMCTS): class GO2CfgREMCTS(LeggedRobotCfgREMCTS):
@@ -320,5 +320,5 @@ class GO2CfgREMCTS(LeggedRobotCfgREMCTS):
class runner(LeggedRobotCfgREMCTS.runner): class runner(LeggedRobotCfgREMCTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_rem_cts' experiment_name = 'go2_rem_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500

View File

@@ -253,7 +253,7 @@ class GO2CfgPPO(LeggedRobotCfgPPO):
class runner(LeggedRobotCfgPPO.runner): class runner(LeggedRobotCfgPPO.runner):
run_name = '' run_name = ''
experiment_name = 'go2_ppo' experiment_name = 'go2_ppo'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgCTS(LeggedRobotCfgCTS): class GO2CfgCTS(LeggedRobotCfgCTS):
@@ -261,7 +261,7 @@ class GO2CfgCTS(LeggedRobotCfgCTS):
num_steps_per_env = 24 num_steps_per_env = 24
run_name = '' run_name = ''
experiment_name = 'go2_cts' experiment_name = 'go2_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class policy(LeggedRobotCfgCTS.policy): class policy(LeggedRobotCfgCTS.policy):
@@ -279,7 +279,7 @@ class GO2CfgMoECTS(LeggedRobotCfgMoECTS):
class runner(LeggedRobotCfgMoECTS.runner): class runner(LeggedRobotCfgMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_moe_cts' experiment_name = 'go2_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS): class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS):
@@ -290,7 +290,7 @@ class GO2CfgMCPCTS(LeggedRobotCfgMCPCTS):
class runner(LeggedRobotCfgMCPCTS.runner): class runner(LeggedRobotCfgMCPCTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_mcp_cts' experiment_name = 'go2_mcp_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS): class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS):
@@ -300,7 +300,7 @@ class GO2CfgACMoECTS(LeggedRobotCfgACMoECTS):
class runner(LeggedRobotCfgACMoECTS.runner): class runner(LeggedRobotCfgACMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_ac_moe_cts' experiment_name = 'go2_ac_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS): class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS):
@@ -310,7 +310,7 @@ class GO2CfgDualMoECTS(LeggedRobotCfgDualMoECTS):
class runner(LeggedRobotCfgDualMoECTS.runner): class runner(LeggedRobotCfgDualMoECTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_dual_moe_cts' experiment_name = 'go2_dual_moe_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500
class GO2CfgREMCTS(LeggedRobotCfgREMCTS): class GO2CfgREMCTS(LeggedRobotCfgREMCTS):
@@ -320,5 +320,5 @@ class GO2CfgREMCTS(LeggedRobotCfgREMCTS):
class runner(LeggedRobotCfgREMCTS.runner): class runner(LeggedRobotCfgREMCTS.runner):
run_name = '' run_name = ''
experiment_name = 'go2_rem_cts' experiment_name = 'go2_rem_cts'
max_iterations = 150000 max_iterations = 120000
save_interval = 500 save_interval = 500

View File

@@ -2,7 +2,7 @@ from setuptools import find_packages
from distutils.core import setup from distutils.core import setup
setup(name='go2_rl_gym', setup(name='go2_rl_gym',
version='0.1.9', version='1.0.0',
author='Wu Tianyang', author='Wu Tianyang',
license="MIT", license="MIT",
packages=find_packages(), packages=find_packages(),