---
library_name: lerobot
pipeline_tag: robotics
license: other
license_name: openmdw1.1-license
license_link: https://openmdw.ai/license/1-1/
base_model: nvidia/Cosmos3-Edge
tags:
  - lerobot
  - cosmos3
  - robotics
  - pretrained
---

# Cosmos3-Edge base for LeRobot

This repository contains the LeRobot policy conversion of the NVIDIA Cosmos3-Edge base checkpoint. It is a starting point for robot policy post-training. The conversion adds no task-specific fine-tuning.

The policy jointly predicts robot action chunks and future video latents from language and current camera observations, with optional current-state conditioning. It contains the model components needed by the LeRobot policy interface.

| Item | Value |
|---|---|
| Original model developer | NVIDIA |
| Original checkpoint | [nvidia/Cosmos3-Edge](https://huggingface.co/nvidia/Cosmos3-Edge/tree/344d602b128d1bbdacb43b08d0a3626f46343e29) |
| Exact source revision | `344d602b128d1bbdacb43b08d0a3626f46343e29` |
| LeRobot compatibility baseline | `b8f060837ced0a4ce4b136ea65b4d56bc7894bed` |
| Policy type | `cosmos3_edge` |
| Backbone | Nemotron 3 Dense VL MoT |
| Decoder layers / hidden width | 28 / 2048 |
| Stored scalar weights, including frozen components and VAE | 4,563,688,396 |
| Stored tensors | 1188, comprising 1184 BF16 and 4 FP32 |

The stored scalar count is not a count of trainable or active parameters. The package includes a Wan2.2 VAE, tokenizer files, and serialized LeRobot processors.

## Loading and post-training

Use a LeRobot version containing the Cosmos3 policy types. Conversion and CPU loading checks used the exact commit above.

The saved observation and action features are empty. Frame rate and camera size are unresolved. A target dataset must provide these values before constructing the policy. Read the configuration with:

```python
from lerobot.configs.policies import PreTrainedConfig

config = PreTrainedConfig.from_pretrained("HaomingSong/Cosmos3-Edge")
```

The training entry points select different configurations:

- `--policy.path=HaomingSong/Cosmos3-Edge` preserves the saved DROID post-training recipe, including domain 8, 32-step chunks, state conditioning, identity normalization, UniPC with 4 inference steps, and guidance 3. State and action must already use the same layout and width.
- `--policy.type=cosmos3_edge --policy.pretrained_path=HaomingSong/Cosmos3-Edge` uses these weights with the size class's default LIBERO recipe, including domain 5, 16-step chunks, no state conditioning, action quantile normalization, UniPC with 30 steps, and guidance 1. The dataset must match that recipe's camera and action layout.

`load_action_modules` defaults to `true`, preserving the released action modules. For native-style post-training, `--policy.load_action_modules=false` reinitializes the action input projection, action output projection, and action modality embedding. The setting returns to `true` after loading so subsequent reloads preserve trained action modules.

The saved `dtype=float32` supports FP32 master weights during training. Most stored weights retain upstream BF16 precision. The four time-embedder tensors are FP32. For inference set `--policy.dtype=bfloat16`. For evaluation also set `--tf32=false`.

## Scope and evaluation

The LeRobot implementation supports joint video-and-action `wam` mode, a single observation step, and the UniPC sampler. Actual action width comes from the target dataset. This package does not expose the upstream reasoning, audio, or standalone video-generation APIs.

CPU checks verified conversion and strict loading with resolved dataset features. No LIBERO closed-loop success rate is reported for this base conversion. Results for separately fine-tuned LIBERO checkpoints or NVIDIA Policy-DROID models do not measure this checkpoint.

## License and provenance

The NVIDIA model materials use [OpenMDW 1.1](LICENSE). Retain the license and applicable copyright and origin notices when redistributing them. [NOTICE](NOTICE) records the original model and conversion attribution. Separately installed runtime software retains its own licenses.

Consult the original [NVIDIA model card](https://huggingface.co/nvidia/Cosmos3-Edge/blob/344d602b128d1bbdacb43b08d0a3626f46343e29/README.md) and its [safety](https://huggingface.co/nvidia/Cosmos3-Edge/blob/344d602b128d1bbdacb43b08d0a3626f46343e29/SAFETY.md), [bias](https://huggingface.co/nvidia/Cosmos3-Edge/blob/344d602b128d1bbdacb43b08d0a3626f46343e29/BIAS.md), [privacy](https://huggingface.co/nvidia/Cosmos3-Edge/blob/344d602b128d1bbdacb43b08d0a3626f46343e29/PRIVACY.md), and [explainability](https://huggingface.co/nvidia/Cosmos3-Edge/blob/344d602b128d1bbdacb43b08d0a3626f46343e29/EXPLAINABILITY.md) notes for upstream limitations.