Pi0 Pre Train 100

🚀 Model Card for pi0_pre_train_100

This policy is trained and hosted on the Hub via LeRobot. It provides a solution for robotics - related tasks. Check out the full documentation at LeRobot Docs.

🚀 Quick Start

📦 Installation

Since no specific installation steps are provided, we'll skip this section.

💻 Usage Examples

Basic Usage

This section shows how to train a policy from scratch and evaluate it.

Train from scratch

python lerobot/scripts/train.py \
  --dataset.repo_id=<user_or_org>/<dataset> \
  --policy.type=act \
  --output_dir=outputs/train/<desired_policy_repo_id> \
  --job_name=lerobot_training \
  --policy.device=cuda \
  --policy.repo_id=<user_or_org>/<desired_policy_repo_id> \
  --wandb.enable=true

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

Evaluate the policy

python -m lerobot.record \
  --robot.type=so100_follower \
  --dataset.repo_id=<user_or_org>/eval_<dataset> \
  --policy.path=<user_or_org>/<desired_policy_repo_id> \
  --episodes=10

Prefix the dataset repo with eval_ and supply --policy.path pointing to a local or hub checkpoint.

📚 Documentation

For a complete walkthrough, see the training guide.

📄 License

This project is licensed under the Apache - 2.0 license.