Files
lerobot/examples/training/pi052_hirobot.slurm
T
Pepijn ecbac17196 chore(training): align pi052_hirobot.slurm with the operator's actual command
Match the working SmolVLA2 launch pattern so the two SLURM scripts
are interchangeable:

  * literal NUM_PROCESSES / BATCH_SIZE / STEPS (no env-var defaults)
  * STEPS=10000 to match the next SmolVLA2 run
  * save_freq=$STEPS so only the final checkpoint is saved
  * dropouts 0.1/0.1/0.1 (mild — matches the operator's iteration)
  * flow_loss_weight / text_loss_weight come from the PI052Config
    defaults (10.0 / 1.0 per Pi 0.5 paper §IV.D), no need to pass
    them explicitly

Job name and policy_repo_id mirror the SmolVLA2 ``_tool-g2`` naming
so the two runs can be compared side-by-side in WandB.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-13 11:03:09 +02:00

59 lines
2.0 KiB
Bash

#!/bin/bash
#SBATCH --job-name=pi052-hirobot-10k
#SBATCH --partition=hopper-prod
#SBATCH --qos=high
#SBATCH --time=48:00:00
#SBATCH --ntasks=1
#SBATCH --gpus-per-task=8
set -euo pipefail
cd "${LEROBOT_ROOT:-$HOME/lerobot}"
export PATH="$HOME/miniconda3/bin:$HOME/.local/bin:$PATH"
export LD_LIBRARY_PATH="$HOME/miniconda3/lib:${LD_LIBRARY_PATH:-}"
export NCCL_TIMEOUT="${NCCL_TIMEOUT:-1800}"
export HF_HUB_DOWNLOAD_TIMEOUT="${HF_HUB_DOWNLOAD_TIMEOUT:-120}"
export WANDB_INIT_TIMEOUT="${WANDB_INIT_TIMEOUT:-300}"
DATASET="pepijn223/super_poulain_full_tool3"
POLICY_REPO_ID="pepijn223/pi052_hirobot_super_poulain_tool-g2"
JOB_NAME="pi052-hirobot-super-poulain-tool-g2-10k"
NUM_PROCESSES=8
BATCH_SIZE=32
STEPS=10000
RUN_ID="${SLURM_JOB_ID:-$(date +%Y%m%d_%H%M%S)}"
OUTPUT_DIR="/fsx/pepijn/outputs/train/pi052_hirobot_super_poulain_tool3_10k_${RUN_ID}"
echo "Training pi052 on $DATASET with ${NUM_PROCESSES} GPUs, batch size ${BATCH_SIZE}/GPU, ${STEPS} steps"
echo "Output directory: $OUTPUT_DIR"
accelerate launch --multi_gpu --num_processes="$NUM_PROCESSES" \
-m lerobot.scripts.lerobot_train \
--policy.type=pi052 \
--policy.recipe_path=recipes/pi052_hirobot.yaml \
--dataset.repo_id="$DATASET" \
--dataset.revision=main \
--dataset.video_backend=pyav \
--output_dir="$OUTPUT_DIR" \
--job_name="$JOB_NAME" \
--policy.repo_id="$POLICY_REPO_ID" \
--policy.compile_model=false \
--policy.device=cuda \
--policy.tokenizer_max_length=512 \
--steps="$STEPS" \
--policy.scheduler_decay_steps="$STEPS" \
--batch_size="$BATCH_SIZE" \
--wandb.enable=true \
--wandb.disable_artifact=true \
--wandb.project=hirobot \
--log_freq=100 \
--save_freq="$STEPS" \
--num_workers=0 \
--dataset.image_transforms.enable=true \
--dataset.image_transforms.max_num_transforms=3 \
--dataset.image_transforms.random_order=true \
--policy.plan_dropout_prob=0.1 \
--policy.memory_dropout_prob=0.1 \
--policy.subtask_dropout_prob=0.1 \