Fix broken and redirected repo links

This commit is contained in:
RemiFabre
2026-08-28 14:37:54 +02:00
parent 10c600e54e
commit 1261013e7e
3 changed files with 7 additions and 7 deletions
+3 -3
View File
@@ -31,8 +31,8 @@ policies are lazy-loaded on the first switch.
| Crouch (rollers) | `BEST_roller_crouch.onnx` | One-shot crouch-glide: sinks low over ~3.5 s and stands back up (phase-encoded command) |
Policies and MJCF model from
[apirrone/microduck_runtime](https://github.com/apirrone/microduck_runtime)
and [apirrone/mjlab_microduck](https://github.com/apirrone/mjlab_microduck).
[pollen-robotics/microduck](https://github.com/pollen-robotics/microduck)
and [pollen-robotics/microduck_rl](https://github.com/pollen-robotics/microduck_rl).
## Controls
@@ -104,7 +104,7 @@ UI intents in).
official `@mujoco/mujoco` WASM bindings.
- Both variants share the exact same policy interface: 61D observation
(gyro, projected gravity, 14 joint pos/vel, last action, 13D command)
and 14 position-targets, matching `mjlab_microduck/scripts/infer_policy.py`.
and 14 position-targets, matching [`microduck_rl/scripts/infer_policy.py`](https://github.com/pollen-robotics/microduck_rl/blob/main/scripts/infer_policy.py).
The roller variant adds 4 passive wheel hinges that appear in `qpos`
(zeroed in the keyframe) but not in the observation.
- Rendering is a three.js rig built from `kinematics.json` /
+1 -1
View File
@@ -1,5 +1,5 @@
// Gamepad input source (see controller.js for the source interface
// contract). Same mapping as the robot runtime (microduck_runtime):
// contract). Same mapping as the robot runtime (pollen-robotics/microduck):
//
// Left stick vertical = vx (asymmetric fwd/back), horizontal = turn,
// EMA-smoothed like the runtime's cmd_alpha.
+3 -3
View File
@@ -3,10 +3,10 @@
//
// Physics runs in MuJoCo compiled to WebAssembly (the official
// @mujoco/mujoco bindings), stepping the same MJCF the policies were
// trained on (apirrone/mjlab_microduck). The controller is one of the
// exported ONNX checkpoints from apirrone/microduck_runtime, executed with
// trained on (pollen-robotics/microduck_rl). The controller is one of the
// exported ONNX checkpoints from pollen-robotics/microduck, executed with
// onnxruntime-web at 50 Hz (timestep 0.005 s, decimation 4) - exactly the
// loop from mjlab_microduck/scripts/infer_policy.py.
// loop from microduck_rl/scripts/infer_policy.py.
//
// Obs layout (61D, "new-cmd-obs" flavor, from the ONNX metadata):
// [base_ang_vel(3), projected_gravity(3), joint_pos(14), joint_vel(14),