pi05_no_dest_firstandlast5_injection_allenv

Full JAX/orbax checkpoints for a RoboPRO obstacle-attention π₀.₅ fine-tune.

Each step directory contains everything needed to resume training or run eval:

• params/ — model weights
• train_state/ — optimizer / train state
• assets/ — norm_stats.json
• _CHECKPOINT_METADATA — orbax item handlers (params, train_state, assets)
• train_config.py — TrainConfig snippet to paste into src/openpi/training/config.py (same pattern as mzxuan/robopro_jax_30000)