videron/pi05_lego_maximus_aug_30k

This is a Lerobot 0.4.4 Model, it will not perform well when using 0.5.1

/usr/local/bin/lerobot-train --dataset.repo_id=staudi25/lego_pickup_maximus_plus --job_name=pi05_so101_lego_pickup --policy.type=pi05 --policy.pretrained_path=lerobot/pi05_base --policy.device=cuda --wandb.enable=true --log_freq=1 --steps=30000 --save_freq=5000 --dataset.image_transforms.enable=true --policy.repo_id=staudi25/pi05_lego_maximus_plus_aug_30k --policy.dtype=bfloat16 --output_dir=outputs/train/pi05_lego_maximus_plus_aug_30k --policy.compile_model=true --policy.gradient_checkpointing=true --policy.freeze_vision_encoder=false --policy.train_expert_only=false --batch_size=32 --policy.scheduler_decay_steps=29000 --policy.scheduler_warmup_steps=1000 --num_workers=16

Final Loss .021

https://www.youtube.com/watch?v=VhD7yG08eDY

https://github.com/videron-ai/lerobot-inference