hangX/lerobot
1
0
Fork 0
forked from tangger/lerobot
Code Pull Requests Activity

Change HILSerlRobotEnvConfig to inherit from EnvConfig

Browse Source 
Added support for hil_serl classifier to be trained with train.py
run classifier training by python lerobot/scripts/train.py --policy.type=hilserl_classifier
fixes in find_joint_limits, control_robot, end_effector_control_utils
This commit is contained in:
Michel Aractingi
2025-03-27 10:23:14 +01:00
committed by AdilZouitine
parent 052a4acfc2
commit d0b7690bc0
13 changed files with 389 additions and 340 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
lerobot/scripts/train.py
View File

@@ -122,6 +122,9 @@ def make_optimizer_and_scheduler(cfg, policy):
optimizer = VQBeTOptimizer(policy, cfg)
lr_scheduler = VQBeTScheduler(optimizer, cfg)
elif cfg.policy.name == "hilserl_classifier":
optimizer = torch.optim.AdamW(policy.parameters(), cfg.policy.learning_rate)
lr_scheduler = None
else:
raise NotImplementedError()