lerobot so-arm101复现步骤

地址： https://huggingface.co/docs/lerobot/so101?assembly=Leader

一、飞特舵机选型

 分为标准版和大扭矩版标准版和大扭矩版主臂舵机相同：1.sts3215-c044   减速比1：1912.sts3215-c001   减速比1：3453.sts3215-c044   减速比1：1914.sts3215-c046   减速比1：1475.sts3215-c046   减速比1：1476.sts3215-c046   减速比1：147标准版从臂舵机：sts3215-c001   减速比1：345**大扭矩版从臂舵机：sts3215-c018   减速比1：345**标准版：1128元    大扭矩版1178元

二、结构件打印

结构件： 160元

三、其他物料

1.微雪舵机驱动板2个，53元
2.USB-C线2根
3.电源2个，7V4A+12V6A
4.木工夹4个
5.螺丝：2个

四、常用命令

sudo chmod 666 /dev/ttyACM*

lerobot-teleoperate
–robot.type=so101_follower
–robot.port=/dev/ttyACM0
–robot.id=my_awesome_follower_arm
–robot.cameras=“{ front: {type: opencv, index_or_path: 10, width: 640, height: 480, fps: 30}, side: {type: opencv, index_or_path: 8, width: 640, height: 480, fps: 30}}”
–teleop.type=so101_leader
–teleop.port=/dev/ttyACM1
–teleop.id=my_awesome_leader_arm
–display_data=true

lerobot-record
–robot.type=so101_follower
–robot.port=/dev/ttyACM0
–robot.id=my_awesome_follower_arm
–robot.cameras=“{ front: {type: opencv, index_or_path: 10, width: 640, height: 480, fps: 30}, side: {type: opencv, index_or_path: 8, width: 640, height: 480, fps: 30}}”
–teleop.type=so101_leader
–teleop.port=/dev/ttyACM1
–teleop.id=my_awesome_leader_arm
–display_data=true
–dataset.repo_id=seeedstudio123/test
–dataset.single_task=“Grab the black cube”
–dataset.push_to_hub=false

--------------------------------------------------------------训练[act,diffusion,pi0,pi0fast,sac,smolvla]-------------------------------------------------------

lerobot-train
–dataset.repo_id=seeedstudio123/test
–policy.type=act
–output_dir=outputs/train/act_so101_test
–job_name=act_so101_test
–policy.device=cuda
–wandb.enable=false
–policy.push_to_hub=false
–steps=300000

lerobot-train
–dataset.repo_id=seeedstudio123/test
–policy.type=smolvla
–output_dir=outputs/train/smolvla_so101_test
–job_name=smolvla_so101_test
–policy.device=cuda
–wandb.enable=false
–policy.push_to_hub=false
–steps=300000

lerobot-train
–dataset.repo_id=seeedstudio123/test
–policy.type=smolvla
–output_dir=outputs/train/smolvla_so101_test
–job_name=smolvla_so101_test
–policy.device=cuda
–wandb.enable=false
–policy.push_to_hub=false
–steps=300000

lerobot-train
–dataset.repo_id=seeedstudio123/test
–policy.type=sac
–output_dir=outputs/train/sac_so101_test
–job_name=sac_so101_test
–policy.device=cuda
–wandb.enable=false
–policy.push_to_hub=false
–steps=300000

-----------------------------------------------------------------------------------------推理-------------------------------------------------------------------------------------

lerobot-record
–robot.type=so101_follower
–robot.port=/dev/ttyACM0
–robot.cameras=“{ front: {type: opencv, index_or_path: 10, width: 640, height: 480, fps: 30}, side: {type: opencv, index_or_path: 6, width: 640, height: 480, fps: 30}}”
–robot.id=my_awesome_follower_arm
–display_data=false
–dataset.repo_id=seeed/eval_test123
–dataset.single_task=“Put lego brick into the transparent box”
–policy.path=outputs/train/act_so101_test/checkpoints/last/pretrained_model

lerobot-record
–robot.type=so101_follower
–robot.port=/dev/ttyACM0
–robot.cameras=“{ front: {type: opencv, index_or_path: 10, width: 640, height: 480, fps: 30}, side: {type: opencv, index_or_path: 6, width: 640, height: 480, fps: 30}}”
–robot.id=my_awesome_follower_arm
–display_data=false
–dataset.repo_id=seeed/eval_diffusiontest123
–dataset.single_task=“Put lego brick into the transparent box”
–policy.path=outputs/train/diffusion_so101_test/checkpoints/last/pretrained_model

lerobot-record
–robot.type=so101_follower
–robot.port=/dev/ttyACM0
–robot.cameras=“{ front: {type: opencv, index_or_path: 10, width: 640, height: 480, fps: 30}, side: {type: opencv, index_or_path: 6, width: 640, height: 480, fps: 30}}”
–robot.id=my_awesome_follower_arm
–display_data=false
–dataset.repo_id=seeed/eval_smolvlatest123
–dataset.single_task=“Put lego brick into the transparent box”
–policy.path=outputs/train/smolvla_so101_test/checkpoints/last/pretrained_model

--------------------------------------------------------------------------问题点-----------------------------------------------------------------------------------------------

INFO 2025-10-25 17:31:27 t_record.py:341 No policy or teleoperator provided, skipping action generation.This is likely to happen when resetting the environment without a teleop device.The robot won’t be at its rest position at the start of the next episode.

lerobot-train
–config_path=outputs/train/smolvla_so101_test/checkpoints/last/pretrained_model/train_config.json
–resume=true