强化学习第一版

This commit is contained in:
2026-09-11 16:56:39 +08:00
parent 2bc4c7ed82
commit 4699b0d45f
78 changed files with 27978 additions and 706 deletions
+15
View File
@@ -0,0 +1,15 @@
{
"problem": 4,
"timesteps": 40000,
"seed": 20260911,
"device": "cuda",
"directional_probability": 1.0,
"wall_seconds": 235.8,
"model": "output\\rl\\q4_ppo.zip",
"init": "output/rl/q4_ppo_bc_init.zip",
"episodes": 40,
"success_rate": 0.9,
"mean_steps": 4.45,
"mean_episode_reward": 12.621703708323922,
"successes": 36
}