8x54zj-m
|
00cb6f9857
|
Tune DreamWaQ gait rewards
|
2026-07-22 02:22:04 +08:00 |
|
8x54zj-m
|
3648551043
|
Backup DreamWaQ rslrl stability fixes
|
2026-07-22 02:17:01 +08:00 |
|
8x54zj-m
|
c664e7422f
|
6level，但是表现一般
|
2026-07-02 10:53:17 +08:00 |
|
8x54zj-m
|
482dabb893
|
微调参数
|
2026-07-01 21:02:40 +08:00 |
|
8x54zj-m
|
63b235d186
|
fix: MuJoCo sim2sim default_angles match MotrixSim training
|
2026-07-01 20:55:16 +08:00 |
|
8x54zj-m
|
2bd4c05e1b
|
fix: fixed schedule, init_noise 0.5, AdaBoot cap 0.9 (keep vel head gradient)
|
2026-07-01 16:56:07 +08:00 |
|
8x54zj-m
|
a3562fbab1
|
fix: storage dones/rewards shape squeeze, NaN guard in AdaBoot
|
2026-07-01 16:02:21 +08:00 |
|
8x54zj-m
|
be193199e5
|
feat: AdaBoot — paper CV of episodic reward (not velocity error)
|
2026-07-01 15:57:00 +08:00 |
|
8x54zj-m
|
5cbc8e2ca4
|
feat: AdaBoot — adaptive velocity bootstrapping (paper Section II-C)
|
2026-07-01 15:44:30 +08:00 |
|
8x54zj-m
|
308f2090a5
|
fix: VAE predicts next frame (paper: LVAE = MSE(õ_{t+1}, o_{t+1})), history excludes current obs
|
2026-07-01 15:39:56 +08:00 |
|
8x54zj-m
|
3038aa7ac7
|
fix: critic called once (was 3x), adaptive KL schedule restored, upgrade 3m
|
2026-07-01 15:05:50 +08:00 |
|
8x54zj-m
|
ef3ccb2e1a
|
fix: NaN protection — dof nan_to_num, physics reset on crash
|
2026-07-01 10:51:36 +08:00 |
|
8x54zj-m
|
35a6f66511
|
fix: NaN chain break — rewards, std, obs, physics all protected
|
2026-07-01 10:31:47 +08:00 |
|
8x54zj-m
|
1eede03ef0
|
fix: dual-shuffle bug in terrain_origins — robots now spawn on correct terrain columns
|
2026-06-30 22:16:29 +08:00 |
|
8x54zj-m
|
cb62f35ef6
|
fix: boundary termination, physics crash handler, action_scale 0.25, play script update
|
2026-06-30 19:03:57 +08:00 |
|
8x54zj-m
|
7affccc695
|
fix: noise scales *obs_scale (match upstream), noise_joint_angle 0.03→0.01
|
2026-06-30 17:30:24 +08:00 |
|
8x54zj-m
|
964b409aa1
|
fix: action_scale 0.05→0.25 (upstream value, 5x difference)
|
2026-06-30 17:29:16 +08:00 |
|
8x54zj-m
|
a5e5d5029e
|
fix: dt in state.reward, feet_air_time first_contact timing
|
2026-06-30 16:42:48 +08:00 |
|
8x54zj-m
|
361502a8a5
|
fix: terrain start level 0, clear ep_report after consume
|
2026-06-30 15:43:02 +08:00 |
|
8x54zj-m
|
6348cf72bc
|
fix: squeeze old_log_prob and advantages dims (match upstream)
|
2026-06-30 15:39:03 +08:00 |
|
8x54zj-m
|
aa17ea853e
|
fix: VAE loss /num_mini_batches, KL use raw sum (align with upstream)
|
2026-06-30 15:38:26 +08:00 |
|
8x54zj-m
|
b0f1da4596
|
fix: clamp std before distribution, lower init_noise to 0.5, NaN guard
|
2026-06-30 15:35:58 +08:00 |
|
8x54zj-m
|
f2a8e0e2ff
|
fix: remove broken feet_air_time override, use base class
|
2026-06-30 15:03:52 +08:00 |
|
8x54zj-m
|
3fd2793f28
|
fix: register dreamwaq env, disable missing stair envs
|
2026-06-30 14:56:01 +08:00 |
|
8x54zj-m
|
422421d263
|
feat: DreamWaQ full replication — env, terrain, CENet, PPO
|
2026-06-30 14:53:12 +08:00 |
|
motphys-developers
|
c569f4d6d9
|
Merge pull request #5 from Motphys/docs/add-motrixlab-citation
docs: add MotrixLab citation metadata and README sections
|
2026-05-25 16:46:11 +08:00 |
|
motphys-developers
|
e1421d1055
|
chore: release v0.3.0
|
2026-04-02 03:45:10 +00:00 |
|
motphys-developers
|
c84d382b8c
|
chore: release v0.2.0.post1
|
2026-02-10 16:53:35 +08:00 |
|
motphys-developers
|
b568ac5600
|
chore: release v0.2.0
|
2026-02-10 08:08:11 +00:00 |
|
jyf23
|
dbfa9e31fa
|
add go2 training
|
2026-01-08 15:24:11 +08:00 |
|
motphys-developers
|
62011bb24f
|
chore: release v0.1.0
(cherry picked from commit 82525f882f3924a332d9ce40bf64255d0d14f6a4)
|
2026-01-08 15:07:57 +08:00 |
|
motphys-developers
|
13cfbce9a7
|
chore: release v0.0.2
|
2025-11-21 08:51:49 +00:00 |
|
motphys-developers
|
5133830b5a
|
chore: release v0.0.1
|
2025-11-20 17:39:31 +08:00 |
|