oc2_sushi_1_3

sushi 1-3

Evaluated Unknown · Uploaded Unknown by Rowan G. Last edited by Rowan G.

agent_0

Observation and actions

{
  "actions": {
    "action_dropout": 0.0,
    "action_selection": "sampled",
    "random_action_prob": 0.0
  },
  "observation": {
    "agent_view_size": null,
    "shape": [
      10,
      14,
      48
    ]
  }
}

agent_1

imitation learning maybe · 1.5K updates

Observation and actions

{
  "actions": {
    "action_dropout": 0.0,
    "action_selection": "sampled",
    "random_action_prob": 0.0
  },
  "observation": {
    "agent_view_size": null,
    "shape": [
      10,
      14,
      48
    ]
  }
}

Statistics across all 1 episodes

episodes: 1
episodes with plate washing: 1
episodes with washed plate reuse: 1
finite predictions: Yes
mean correct deliveries: 7
mean first delivery step among successes: 78
mean raw team reward: 140
mean second delivery step among successes: 140
status: complete
success rate at least one: 1
success rate at least two: 1
total correct deliveries: 7
total dropout selections: 0
total perturbed actions: 0
total random action selections: 0

Selected episode

Select an episode to inspect its recording.

Evaluation protocol and metric definitions

{
  "action_dropout": 0.0,
  "action_selection": "sampled",
  "algorithm_source_sha256": "aac7932be5e7ce08375e30c3b20d78c44d0ec3d9c25a9b0be1b39c7537dde7b1",
  "backend": "cpu",
  "environment_source_sha256": "ffe3196e464d8b0a570fcdb9f3d4f8ff5d2946438f9230315954d1e8d2b926ad",
  "episode_seeds": [
    6000
  ],
  "episode_steps": 600,
  "evaluator_source_sha256": "c4081f1c8a2d7fbfee6c2b1e42adbe7eb7b6cdcdb7a465fd19471c78ff3100f5",
  "expert_or_planner_used": false,
  "greedy_scope": "Greedy repetitions use the identical deterministic initial situation and are not independent starting-state tests.",
  "initial_conditions": [
    {
      "facings": [
        3,
        3
      ],
      "positions": [
        [
          4,
          3
        ],
        [
          11,
          3
        ]
      ],
      "seed": 6000
    }
  ],
  "initial_state": "standard empty kitchen; fixed chef starts",
  "layout": "oc2_sushi_1_3",
  "learning_during_evaluation": false,
  "observation_shape": [
    10,
    14,
    48
  ],
  "perturbation_rng": "Independent folded streams from eval seed; original policy/reset key chain is unchanged.",
  "perturbation_scope": "Independent per-chef/per-step exogenous action overrides; initial jitter changes only valid chef poses on this fixed layout.",
  "policy_sha256": "8d086d8bd3f18e751e08974145dae825f0e18ed4b1edaed93515c7b4e474c46a",
  "random_action_prob": 0.0,
  "random_facing": false,
  "recording_coverage": "all_episodes_full_resolution",
  "replay_episode": null,
  "reward": "raw shared environment reward counted once; no training shaping",
  "seed_scope": "Reset ignores seeds. Seeds vary policy action sampling only; no unseen layouts or randomized starts.",
  "spawn_jitter": 0,
  "spawn_jitter_distance": "Manhattan",
  "state_action_convention": "steps[0] initial; steps[t] after attached action t",
  "training_episode_steps": 600,
  "upstream_revision": "976aeb152cb184a5095021968bba94da96eb6394"
}

Effective environment configuration

{
  "action_meanings": {
    "0": "RIGHT",
    "1": "DOWN",
    "2": "LEFT",
    "3": "UP",
    "4": "STAY",
    "5": "PICKUP_PLACE",
    "6": "WORK"
  },
  "config": {
    "agent_view_size": null,
    "layout": "oc2_sushi_1_3",
    "rules": {
      "max_steps": 600
    }
  },
  "layout": "oc2_sushi_1_3",
  "name": "overcooked_v2",
  "renderer": {
    "chop_ticks": 6,
    "cook_ticks": 48,
    "kind": "sushi",
    "wash_ticks": 12
  }
}

Evaluation code and dependency provenance

{
  "git_commit": "b927e547df45f18e1ab3a182d4fa6ac3f0b11166",
  "git_dirty": true,
  "sources": {
    "algorithm_sha256": "aac7932be5e7ce08375e30c3b20d78c44d0ec3d9c25a9b0be1b39c7537dde7b1",
    "algorithm_source_sha256": "aac7932be5e7ce08375e30c3b20d78c44d0ec3d9c25a9b0be1b39c7537dde7b1",
    "environment_sha256": "ffe3196e464d8b0a570fcdb9f3d4f8ff5d2946438f9230315954d1e8d2b926ad",
    "environment_source_sha256": "ffe3196e464d8b0a570fcdb9f3d4f8ff5d2946438f9230315954d1e8d2b926ad",
    "evaluation_core_sha256": "858be07bd210038cd666eaf222e16328bc97dbbe5ed0857167d8857e0e262a5a",
    "evaluator_source_sha256": "c4081f1c8a2d7fbfee6c2b1e42adbe7eb7b6cdcdb7a465fd19471c78ff3100f5"
  },
  "training_checkpoint_provenance": {
    "run:f4b9a13d-e19d-4eda-896c-6a2f7479861d:sha256:8d086d8bd3f18e751e08974145dae825f0e18ed4b1edaed93515c7b4e474c46a": {
      "training_resumes": []
    }
  },
  "upstream_revision": "976aeb152cb184a5095021968bba94da96eb6394"
}

Evaluation device (separate from training device)

{
  "architecture": "x86_64",
  "backend": "cpu",
  "devices": [
    {
      "kind": "cpu",
      "platform": "cpu"
    }
  ],
  "logical_cpu_count": 16,
  "os": "Linux",
  "os_release": "6.18.40.1-microsoft-standard-WSL2",
  "python": "3.12.3",
  "versions": {
    "flax": "0.10.2",
    "jax": "0.4.38",
    "jaxlib": "0.4.38",
    "numpy": "1.26.4",
    "optax": "0.2.4"
  }
}