AdrianLlopart's picture
docs: HF model card β€” best-of-both front-matter derived from manifest
946a7db verified
|
Raw
History Blame Contribute Delete
9.62 kB
---
language:
- en
license: apache-2.0
pipeline_tag: robotics
tags:
- OpenRAL
- rskill
- molmoact2
- vision-language-action
- nf4
- 4-bit
- so100_follower
- so101_follower
- transformers
- vla
- so101
- so100
- manipulation
base_model:
- allenai/MolmoAct2-SO100_101
base_model_relation: quantized
inference: false
---
# rskill-molmoact2-so101-nf4
> **OpenRAL rSkill** β€” MolmoAct2 (Ai2's open action reasoning model: a
> Molmo2-ER embodied-reasoning VLM backbone with a flow-matching
> continuous-action expert) finetuned on the
> [SO-100/SO-101](https://huggingface.co/allenai/MolmoAct2-SO100_101) teleop
> mixture and NF4-quantized so the ~5.5 B-param model fits an 8 GB GPU.
> Robots: SO-100 and SO-101 follower arms. **Apache-2.0 weights** β€” commercial
> use permitted.
This package wraps `hf://OpenRAL/rskill-molmoact2-so101-nf4` (an
NF4-quantized mirror of `allenai/MolmoAct2-SO100_101`) with a `rskill.yaml`
manifest that adds capability checking, license surfacing, latency budgets,
and local registry integration. It does **not** copy model weights β€” they
live on the Hub.
> **Required sim config knob:** this checkpoint uses normalization statistics
> tagged `"so100_so101_molmoact2"`. Any `SimEnvironment` config that drives
> this rSkill must set `vla.extra.norm_tag: "so100_so101_molmoact2"` β€”
> omitting it silently applies the adapter's default `"libero"` norm stats and
> produces garbage actions.
## What this skill does
Performs tabletop manipulation β€” picking, placing, grasping, and transporting
objects β€” on the SO-100 and SO-101 follower arms. The MolmoAct2 backbone
reasons about the scene in 3D and the flow-matching action expert emits a
continuous absolute joint-position action chunk that the adapter replays one
step at a time.
| Field | Value |
| --- | --- |
| Actions | pick, place, pick_and_place, grasp |
| Objects | diverse tabletop objects |
| Scenes | tabletop |
| Embodiments | `so100_follower`, `so101_follower` |
## How it works
MolmoAct2 grafts a modern DiT-style flow-matching continuous-action expert
onto the Molmo2-ER discrete-token VLM via per-layer KV-cache conditioning
(arXiv:2605.02881). It ships as a transformers **custom-code** model
(`trust_remote_code`, `auto_map` β†’ `MolmoAct2ForConditionalGeneration`), not a
lerobot policy. The OpenRAL `molmoact2` adapter
(`python/sim/src/openral_sim/policies/molmoact2.py`) loads it via
`AutoModelForImageTextToText.from_pretrained` + `AutoProcessor` from the
manifest's `source_repo`, NF4-quantizes every Linear with β‰₯4M weight elements
via bitsandbytes, overlays the prequantized pack from `weights_uri`, then drives
it through the checkpoint's own `predict_action(...)` continuous-action API. Two
RGB camera streams plus a 6-D proprio state go in; a `(chunk_size, 6)` absolute
joint-position chunk comes out, replayed one step at a time and re-inferred
when the queue empties.
The adapter reads `norm_tag` from `vla.extra.norm_tag`; this rSkill requires
`"so100_so101_molmoact2"` β€” set it explicitly in every `SimEnvironment` config.
### Observation β†’ action contract
| Direction | Key | Shape | Notes |
| --- | --- | --- | --- |
| in | `observation.images.camera1` | `(1, 3, H, W) float32 [0,1]` | overhead view (β†’ model `top`) |
| in | `observation.images.camera2` | `(1, 3, H, W) float32 [0,1]` | wrist/side view (β†’ model `side`) |
| in | `observation.state` | `(1, 6)` float32 | SO-101 6-D joint positions (rad) |
| out | action chunk | `(10, 6)` float32 | absolute joint-position targets |
**Camera aliases (for `so101_box` scene):** `oak_top β†’ top`, `wrist β†’ side`.
Override per-scene via `vla.extra` if your scene uses different camera names.
## Upstream model / training
The wrapped weights come from Ai2's `allenai/MolmoAct2-SO100_101` checkpoint β€”
the base `allenai/MolmoAct2` foundation model finetuned on the SO-100/SO-101
teleop dataset mixture with absolute joint-pose control and annotated language
instructions. This rSkill repackages an NF4-quantized mirror of those weights;
it does **not** retrain or copy the full-precision weights.
| Field | Value |
| --- | --- |
| Source repo | [`allenai/MolmoAct2-SO100_101`](https://huggingface.co/allenai/MolmoAct2-SO100_101) |
| Base model | [`allenai/MolmoAct2`](https://huggingface.co/allenai/MolmoAct2) |
| Paper | [arxiv:2605.02881](https://arxiv.org/abs/2605.02881) β€” *MolmoAct2: Action Reasoning Models for Real-world Deployment* |
| License | apache-2.0 (code + weights) |
| Parameters | ~5.5 B |
| Training data | SO-100/SO-101 teleop mixture (absolute joint-pose, annotated language) |
| norm_tag | `"so100_so101_molmoact2"` β€” **required** in `vla.extra.norm_tag` |
## Supported robots
| Robot | Embodiment tag | Status | Notes |
| --- | --- | --- | --- |
| SO-101 follower | `so101_follower` | ⚑ experimental | Native training embodiment; numbers not yet locally reproduced. |
| SO-100 follower | `so100_follower` | ⚑ experimental | Shares identical 6-DoF kinematics; covered by training mixture. |
## Sensors required
| Key | Modality | Min resolution | Format |
| --- | --- | --- | --- |
| `observation.images.camera1` | RGB | 224 Γ— 224 | `float32` |
| `observation.images.camera2` | RGB | 224 Γ— 224 | `float32` |
| `observation.state` | proprioception | (6,) | `float32` |
## Manifest summary
| Field | Value |
| --- | --- |
| `name` | `OpenRAL/rskill-molmoact2-so101-nf4` |
| `version` | `0.1.0` |
| `license` | `apache-2.0` |
| `role` | `s1` |
| `embodiment_tags` | `["so100_follower", "so101_follower"]` |
| `runtime` / `quantization.dtype` | `pytorch` / `int4` (NF4) |
| `weights_uri` | `hf://OpenRAL/rskill-molmoact2-so101-nf4` |
| `chunk_size` / `n_action_steps` | 10 / 10 (full chunk replay) |
| `latency_budget.per_chunk_ms` | 1000 ms |
| `commercial_use_allowed` | `true` (Apache-2.0) |
| `image_preprocessing.image_max_crops` | `4` (secondary vision lever; processor default is 8 β€” see Memory note) |
| **`norm_tag` (vla.extra)** | **`"so100_so101_molmoact2"` β€” required** |
Full schema: [`openral_core.schemas.RSkillManifest`](../../python/core/src/openral_core/schemas.py).
## Quick start
```python
from openral_rskill.loader import rSkill
pkg = rSkill.from_yaml("rskills/molmoact2-so101-nf4/rskill.yaml")
print(pkg.manifest.name, pkg.manifest.version)
```
```bash
# CLI:
uv run openral rskill install OpenRAL/rskill-molmoact2-so101-nf4
uv run openral rskill check # does this host meet the requirements?
```
### Sim config snippet
```yaml
vla:
id: molmoact2
weights_uri: rskills/molmoact2-so101-nf4
extra:
norm_tag: "so100_so101_molmoact2" # REQUIRED β€” default "libero" is wrong for this checkpoint
# image_max_crops: 6 # optional secondary lever; manifest pins 4 (see note)
```
> **Memory note (measured on an 8 GiB RTX 4070, transformers 5.x).** NF4 makes
> the model ~6.0 GiB resident (the bf16 vocab embeddings + vision tower
> dominate; the nf4 Linears are ~3.5 GiB) and it peaks **~7.63 GiB** during a
> chunk β€” right at the edge of an 8 GiB card (which exposes only ~7.6 GiB
> usable). The decisive enabler is the **CUDA expandable-segments allocator**:
> without it the first forward's ~1.5 GiB embedding `cat` cannot be placed
> contiguously and OOMs. The molmoact2 adapter turns this on automatically
> (`PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True`, via
> `_enable_expandable_segments`) before its first CUDA allocation; export it
> yourself if other GPU work in the process allocates before the policy loads.
> `image_max_crops` (pinned to 4 here) is a *secondary* lever β€” it bounds the
> vision tile count but does **not** by itself decide the 8 GiB fit on these
> checkpoints, and transformers 5.x's fast image processor largely ignores it.
> Leave ~0.4 GiB of headroom: don't run other GPU processes alongside it.
## Reproduction
```bash
just bootstrap && uv sync --all-packages
# Closed-loop rollout against the SO-101 box scene (NF4 weights fit an 8 GB GPU):
openral sim run --config scenes/sim/so101_tube_insertion.yaml \
--rskill rskills/molmoact2-so101-nf4 \
--vla.extra.norm_tag so100_so101_molmoact2
```
Producing / refreshing the NF4 weights on the Hub (one-shot, needs a CUDA
host):
```bash
HF_TOKEN=<write-token> uv run python tools/quantize_rskill.py \
--source allenai/MolmoAct2-SO100_101 \
--target OpenRAL/rskill-molmoact2-so101-nf4 \
--loader transformers --trust-remote-code
```
## Evaluation
`eval/so101.json::status` is **pending** β€” no locally-reproduced benchmark
numbers are available yet. Run the reproduction command in
`eval/so101.json::source.reproduction_cli` to populate.
## License
This rSkill package (`rskill.yaml`, `README.md`, `eval/so101.json`) is
**Apache-2.0**. The wrapped weights at
`hf://OpenRAL/rskill-molmoact2-so101-nf4` (NF4 mirror of
`allenai/MolmoAct2-SO100_101`) are also released under **Apache-2.0** by Ai2 β€”
commercial use is permitted; review the upstream LICENSE before deployment.
## See also
- [`robots/so101_follower/README.md`](../../robots/so101_follower/README.md) β€” RobotDescription manifest.
- [`robots/so100_follower/README.md`](../../robots/so100_follower/README.md) β€” SO-100 variant.
- [`scenes/sim/so101_tube_insertion.yaml`](../../scenes/sim/so101_tube_insertion.yaml) β€” SO-101 sim scene config.
- [`rskills/molmoact2-libero-nf4/README.md`](../molmoact2-libero-nf4/README.md) β€” MolmoAct2 LIBERO variant (Franka Panda).
- [CLAUDE.md Β§6.4](../../CLAUDE.md) β€” rSkill packaging contract.