Back to Lerobot

For reference on model card metadata, see the spec: https://github.com/huggingface/hub-docs/blob/main/modelcard.md?plain=1

src/lerobot/templates/lerobot_rewardmodel_modelcard_template.md

0.6.12.5 KB
Original Source

Reward Model Card for {{ model_name | default("Reward Model ID", true) }}

<!-- Provide a quick summary of what the reward model is/does. -->

{% if model_name == "reward_classifier" %} A reward classifier is a lightweight neural network that scores observations or trajectories for task success, providing a learned reward signal or offline evaluation when explicit rewards are unavailable. {% elif model_name == "sarm" %} A Success-Aware Reward Model (SARM) predicts a dense reward signal from observations, typically used downstream for reinforcement learning or human-in-the-loop fine-tuning when task success is not directly observable. {% elif model_name == "robometer" %} ROBOMETER is a general-purpose video-language robotic reward model built on a fine-tuned Qwen3-VL-4B backbone with progress, preference, and success heads. Given a trajectory video and a task description, it predicts dense, frame-level task progress in [0, 1] and frame-level success probabilities for downstream robot learning, including offline RL, online RL, data filtering and retrieval, and automated failure detection. {% elif model_name == "topreward" %} TOPReward is a zero-shot reward model that extracts token log-probabilities from an off-the-shelf vision-language model (default Qwen3-VL) as a reward signal. Given a video trajectory and a task instruction, it returns the VLM's log-likelihood of the instruction being true, with no fine-tuning required. {% else %} Reward model type not recognized — please update this template. {% endif %}

This reward model has been trained and pushed to the Hub using LeRobot. See the full documentation at LeRobot Docs.


How to Get Started with the Reward Model

Train from scratch

bash
lerobot-train \
  --dataset.repo_id=${HF_USER}/<dataset> \
  --reward_model.type={{ model_name | default("reward_classifier", true) }} \
  --output_dir=outputs/train/<desired_reward_model_repo_id> \
  --job_name=lerobot_reward_training \
  --reward_model.device=cuda \
  --reward_model.repo_id=${HF_USER}/<desired_reward_model_repo_id> \
  --wandb.enable=true

Writes checkpoints to outputs/train/<desired_reward_model_repo_id>/checkpoints/.

Load the reward model in Python

python
from lerobot.rewards import make_reward_model

reward_model = make_reward_model(pretrained_path="<hf_user>/<reward_model_repo_id>")
reward = reward_model.compute_reward(batch)

Model Details

  • License: {{ license | default("[More Information Needed]", true) }}