coord_ring seed 373

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

Training information

Algorithm: ippo_rnn
Seed: 373
Progress unit: steps

The highest uploaded checkpoint may not be the final progress of the local run. Initialization-only runs do not indicate training.

Uploaded checkpoints

coord_ring seed 373 · 9.96M steps

Related evaluations

coord_ring separately trained policies

Checkpoint evidence

[
  {
    "checkpoint_id": "839294b2-f22f-486b-a298-860e1666eac7",
    "id": "839294b2-f22f-486b-a298-860e1666eac7",
    "label": "coord_ring seed 373 \u00b7 9.96M steps",
    "progress": 9961472,
    "progress_unit": "environment_steps",
    "run_id": "84806abf-a649-4964-b5a1-93b31804b490",
    "run_name": "coord_ring seed 373",
    "sha256": "608643a3d137c94bfb122b67432c6f3459e2c604bbe7ef60dade0131fea171e8",
    "source_checkpoint_id": "run:5f3ad928-0796-4d95-9b4a-52b7b2a3714d:sha256:608643a3d137c94bfb122b67432c6f3459e2c604bbe7ef60dade0131fea171e8",
    "training_run_name": "coord_ring seed 373"
  }
]

Training configuration

{
  "ACTIVATION": "relu",
  "ANNEAL_LR": true,
  "CLIP_EPS": 0.2,
  "CNN_FEATURES": 32,
  "ENT_COEF": 0.01,
  "ENV_KWARGS": {
    "agent_view_size": 2,
    "indicate_successful_delivery": false,
    "layout": "coord_ring",
    "max_steps": 400,
    "negative_rewards": true,
    "observation_type": "default",
    "random_agent_positions": true,
    "random_reset": false,
    "sample_recipe_on_delivery": false,
    "start_cooking_interaction": false
  },
  "ENV_NAME": "overcooked_v2",
  "FC_DIM_SIZE": 128,
  "GAE_LAMBDA": 0.95,
  "GAMMA": 0.99,
  "GRU_HIDDEN_DIM": 128,
  "LR": 0.0005,
  "LR_WARMUP": 0.05,
  "MAX_GRAD_NORM": 0.25,
  "MINIBATCH_SIZE": 2048,
  "NUM_ENVS": 256,
  "NUM_MINIBATCHES": 64,
  "NUM_SEEDS": 1,
  "NUM_STEPS": 256,
  "REW_SHAPING_HORIZON": 5000000,
  "SCALE_CLIP_EPS": false,
  "SEED": 373,
  "TOTAL_TIMESTEPS": 10000000,
  "UPDATE_EPOCHS": 4,
  "VF_COEF": 0.5,
  "WANDB_MODE": "disabled"
}

Training code and dependency provenance

{
  "backend": "gpu",
  "upstream_revision": "976aeb152cb184a5095021968bba94da96eb6394",
  "versions": {
    "flax": "0.10.2",
    "jax": "0.4.38",
    "jaxlib": "0.4.38",
    "optax": "0.2.4"
  }
}

Training device

{}