feat: add Go1 training and MuJoCo deployment

deploy/deploy_mujoco/configs/go1.yaml

@@ -0,0 +1,43 @@
+policy_path: "{ROOT_DIR}/deploy/pre_train/go1/policy.pt"
+xml_path: "{ROOT_DIR}/resources/go1/stairs_and_slope.xml"
+
+render_fps: 300
+video_fps: 60
+save_video: false
+
+simulation_duration: 60000000.0
+simulation_dt: 0.002
+control_decimation: 10
+base_init_pos: [0.0, 0.0, 0.34]
+base_init_quat: [1.0, 0.0, 0.0, 0.0]
+
+# Deployment defaults to no extra delay. Training randomizes 0-4 physics steps;
+# this script's delay buffer is measured in 20 ms policy steps instead.
+actuator_delay_min: 0
+actuator_delay_max: 0
+actuator_delay_seed: 0
+
+kps: [28.0, 28.0, 28.0, 28.0, 28.0, 28.0, 28.0, 28.0, 28.0, 28.0, 28.0, 28.0]
+kds: [0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7]
+
+default_angles: [-0.1, 0.8, -1.5, 0.1, 0.8, -1.5,
+                 -0.1, 1.0, -1.5, 0.1, 1.0, -1.5]
+
+lin_vel_scale: 2.0
+ang_vel_scale: 0.25
+dof_pos_scale: 1.0
+dof_vel_scale: 0.05
+action_pos_scale: 0.25
+cmd_scale: [1.0, 1.0, 1.0]
+num_actions: 12
+num_obs: 45
+max_cmd: [1.0, 1.0, 1.0]
+cmd_init: [0.5, 0.0, 0.0]
+history_len: 10
+
+mujoco_joint_names: &joint_names
+  [FR_hip_joint, FR_thigh_joint, FR_calf_joint,
+   FL_hip_joint, FL_thigh_joint, FL_calf_joint,
+   RR_hip_joint, RR_thigh_joint, RR_calf_joint,
+   RL_hip_joint, RL_thigh_joint, RL_calf_joint]
+model_joint_names: *joint_names