publish sprint-2m-reference: perpetual, Mjlab-Sprint2m-MicroDuck

README.md

@@ -0,0 +1,46 @@
+---
+tags:
+- microduck
+- robotics
+- reinforcement-learning
+- onnx
+library_name: onnx
+---
+
+# sprint-2m-reference
+
+Pollen's reference for the Microduck Arena's 2 m Sprint: the sprint_2m challenge of microduck-challenges trained unchanged with its recipe (4096 envs, 3000 iterations, seed 1).
+
+A **perpetual** policy for the [microduck](https://github.com/pollen-robotics/microduck) (61-D observation, 14 actions, 50 Hz). Runs until told otherwise — a gait for the `walk` slot.
+
+## Run it on a robot
+
+```bash
+sudo robotctl policy load walk pollen-robotics/microduck-sprint-2m-reference
+```
+
+The observation normalizer is baked into `policy.onnx`; feed raw observations.
+`manifest.json` follows schema 2 of the microduck policy manifest (`docs/policy-manifest.md` in the daemon repo).
+
+## Training
+
+- **task_id**: `Mjlab-Sprint2m-MicroDuck`
+- **repo**: `https://github.com/pollen-robotics/microduck-challenges.git`
+- **branch**: `main`
+- **commit**: `0808f9872`
+- **checkpoint**: `2999`
+- **seed**: `1`
+- **base**: `mjlab-microduck 0.1.0 @ 981a279c6`
+- **started**: `2026-09-25T22:07:33Z`
+
+## Reproduce
+
+Same code, same `uv.lock`, same command, same seed. Training it again yields a comparable policy, not the same weights: GPU reinforcement learning is not bit-reproducible across machines.
+
+```bash
+git clone https://github.com/pollen-robotics/microduck-challenges.git
+cd microduck-challenges
+git checkout 0808f9872
+uv sync
+uv run train Mjlab-Sprint2m-MicroDuck --env.scene.num-envs 4096 --agent.max-iterations 3000 --agent.seed 1 --agent.logger tensorboard --agent.run-name reference
+```

checkpoint.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bfcd77664e017c5048e05117f40985ecc33abdef46ecd99a01eae82ac7258df
+size 4843445

manifest.json

@@ -0,0 +1,59 @@
+{
+  "schema_version": 2,
+  "model_api": 1,
+  "obs_len": 61,
+  "action_len": 14,
+  "robot": {
+    "model": "microduck",
+    "hw_rev": 1,
+    "servos": "xl330",
+    "control_hz": 50,
+    "accessories": []
+  },
+  "name": "sprint-2m-reference",
+  "kind": "perpetual",
+  "entry_pose": "standing",
+  "description": "Pollen's reference for the Microduck Arena's 2 m Sprint: the sprint_2m challenge of microduck-challenges trained unchanged with its recipe (4096 envs, 3000 iterations, seed 1).",
+  "command": {
+    "encoding": "constant",
+    "idle": [
+      0.0,
+      0.0,
+      0.0
+    ],
+    "twist": "[forward speed m/s, lateral speed m/s, yaw rate rad/s]; the Arena commands [vx, 0, 0]",
+    "head": "unused (zeros)",
+    "body": "unused (zeros)"
+  },
+  "duration_s": null,
+  "slot": "walk",
+  "training": {
+    "repo": "https://github.com/pollen-robotics/microduck-challenges.git",
+    "commit": "0808f9872",
+    "branch": "main",
+    "dirty": false,
+    "command": [
+      "train",
+      "Mjlab-Sprint2m-MicroDuck",
+      "--env.scene.num-envs",
+      "4096",
+      "--agent.max-iterations",
+      "3000",
+      "--agent.seed",
+      "1",
+      "--agent.logger",
+      "tensorboard",
+      "--agent.run-name",
+      "reference"
+    ],
+    "seed": 1,
+    "base": "mjlab-microduck 0.1.0 @ 981a279c6",
+    "started": "2026-09-25T22:07:33Z",
+    "task_id": "Mjlab-Sprint2m-MicroDuck",
+    "checkpoint": 2999,
+    "source_file": "model_2999.pt"
+  },
+  "arena": {
+    "event": "sprint-2m"
+  }
+}

policy.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c93b9fb5d2c0085ac7de13c5f42889057053e06a33beb2b28bf00d74a96776b0
+size 793766