diff --git a/src/preprocess.py b/src/preprocess.py index d451bd3..322eb74 100644 --- a/src/preprocess.py +++ b/src/preprocess.py @@ -2,6 +2,7 @@ from jux.env import JuxEnv from jux.config import JuxBufferConfig from jux.unit import UnitType +from jux.state import State import jax import jax.numpy as jnp @@ -39,20 +40,16 @@ def to_board(x, y, unit_info): return out @jit -def get_unit_feature(unit_mask, unit_type, cargo, power, x, y): +def get_unit_feature(state: State)->jnp.ndarray: ''' - unit_mask : ShapedArray(bool[2, MAX_N_UNITS]) - unit_type : ShapedArray(bool[2, MAX_N_UNITS]) - cargo: ShapedArray(int32[2, MAX_N_UNITS, 4]) - power: ShapedArray(int32[2, MAX_N_UNITS]) - x : ShapedArray(int8[2, MAX_N_UNITS]) - y : ShapedArray(int8[2, MAX_N_UNITS]) - + state: State output: ShapedArray(int8[2, MAP_SIZE, MAP_SIZE, 12]) feature: [light_existence, heavy_existence, (current) ice, ore, water, metal, power, (cargo empty space) ice, ore, water, metal, power] ''' + unit_mask, unit_type, cargo, power, x, y = state.unit_mask, state.units.unit_type, state.units.cargo.stock, state.units.power, state.units.pos.x, state.units.pos.y + light_mask = unit_mask & (unit_type==UnitType.LIGHT) heavy_mask = unit_mask & (unit_type==UnitType.HEAVY) unit_mask_per_type = jnp.stack((light_mask, heavy_mask), axis=-1) @@ -65,9 +62,23 @@ def get_unit_feature(unit_mask, unit_type, cargo, power, x, y): feature = jnp.concatenate((unit_mask_per_type, cargo, power[...,None], cargo_left, battery_left[...,None]), axis=-1) - unit_resource_map = to_board(x, y, feature) + unit_feature_map = to_board(x, y, feature) + + return unit_feature_map + +@jit +def get_factory_feature(state: State, power_previous: jnp.ndarray)->jnp.ndarray: + """ + state: State + output: ShapedArray(int8[2, MAP_SIZE, MAP_SIZE, 7]) + """ + factory_mask, cargo, power, x, y = state.factory_mask, state.factories.cargo.stock, state.factories.power, state.factories.pos.x, state.factories.pos.y + feature = jnp.concatenate((factory_mask[..., None], cargo, power[...,None], power_previous[...,None]), axis=-1) + + factory_feature_map = to_board(x, y, feature) + + return factory_feature_map - return unit_resource_map if __name__=="__main__": diff --git a/src/process_input.ipynb b/src/process_input.ipynb index b3936db..1f7933e 100644 --- a/src/process_input.ipynb +++ b/src/process_input.ipynb @@ -11,9 +11,17 @@ }, { "cell_type": "code", - "execution_count": 27, + "execution_count": 2, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "No GPU/TPU found, falling back to CPU. (Set TF_CPP_MIN_LOG_LEVEL=0 and rerun for more info.)\n" + ] + } + ], "source": [ "import jux\n", "from jux.env import JuxEnv\n", @@ -37,172 +45,785 @@ "\n", "from importlib import reload\n", "\n", - "\n", "MAP_SIZE=64" ] }, { "cell_type": "code", - "execution_count": 22, - "metadata": {}, - "outputs": [], - "source": [ - "reload(ut)\n", - "lux_env, lux_actions = jux.utils.load_replay('replays/52958192.json')\n", - "jux_env, state = JuxEnv.from_lux(lux_env, buf_cfg=JuxBufferConfig(MAX_N_UNITS=200))\n", - "\n", - "state, lux_actions = ut.replay_run_early_phase(jux_env, state, lux_actions)\n", - "state, lux_actions = ut.replay_run_n_late_game_step(100, jux_env, state, lux_actions)" - ] - }, - { - "cell_type": "code", - "execution_count": 4, + "execution_count": 3, "metadata": {}, "outputs": [ { - "data": { - "text/html": [ - "
(\n",
-       "    Board(\n",
-       "        seed=ShapedArray(int32[]),\n",
-       "        factories_per_team=ShapedArray(int8[]),\n",
-       "        map=GameMap(\n",
-       "            rubble=ShapedArray(int8[64,64]),\n",
-       "            ice=ShapedArray(bool[64,64]),\n",
-       "            ore=ShapedArray(bool[64,64]),\n",
-       "            symmetry=ShapedArray(int8[])\n",
-       "        ),\n",
-       "        lichen=ShapedArray(int32[64,64]),\n",
-       "        lichen_strains=ShapedArray(int8[64,64]),\n",
-       "        units_map=ShapedArray(int16[64,64]),\n",
-       "        factory_map=ShapedArray(int8[64,64]),\n",
-       "        factory_occupancy_map=ShapedArray(int8[64,64]),\n",
-       "        factory_pos=ShapedArray(int8[22,2])\n",
-       "    ),\n",
-       "    Unit(\n",
-       "        unit_type=ShapedArray(int8[2,200]),\n",
-       "        action_queue=ActionQueue(\n",
-       "            data=UnitAction(\n",
-       "                action_type=ShapedArray(int8[2,200,20]),\n",
-       "                direction=ShapedArray(int8[2,200,20]),\n",
-       "                resource_type=ShapedArray(int8[2,200,20]),\n",
-       "                amount=ShapedArray(int16[2,200,20]),\n",
-       "                repeat=ShapedArray(int16[2,200,20]),\n",
-       "                n=ShapedArray(int16[2,200,20])\n",
-       "            ),\n",
-       "            front=ShapedArray(int8[2,200]),\n",
-       "            rear=ShapedArray(int8[2,200]),\n",
-       "            count=ShapedArray(int8[2,200])\n",
-       "        ),\n",
-       "        team_id=ShapedArray(int8[2,200]),\n",
-       "        unit_id=ShapedArray(int16[2,200]),\n",
-       "        pos=Position(pos=ShapedArray(int8[2,200,2])),\n",
-       "        cargo=UnitCargo(stock=ShapedArray(int32[2,200,4])),\n",
-       "        power=ShapedArray(int32[2,200])\n",
-       "    ),\n",
-       "    ShapedArray(int16[2000,2]),\n",
-       "    ShapedArray(int16[2]),\n",
-       "    Factory(\n",
-       "        team_id=ShapedArray(int8[2,11]),\n",
-       "        unit_id=ShapedArray(int8[2,11]),\n",
-       "        pos=Position(pos=ShapedArray(int8[2,11,2])),\n",
-       "        power=ShapedArray(int32[2,11]),\n",
-       "        cargo=UnitCargo(stock=ShapedArray(int32[2,11,4]))\n",
-       "    ),\n",
-       "    ShapedArray(int8[22,2]),\n",
-       "    ShapedArray(int8[2]),\n",
-       "    Team(\n",
-       "        team_id=ShapedArray(int8[2]),\n",
-       "        faction=ShapedArray(int8[2]),\n",
-       "        init_water=ShapedArray(int32[2]),\n",
-       "        init_metal=ShapedArray(int32[2]),\n",
-       "        factories_to_place=ShapedArray(int32[2]),\n",
-       "        factory_strains=ShapedArray(int8[2,11]),\n",
-       "        n_factory=ShapedArray(int8[2]),\n",
-       "        bid=ShapedArray(int32[2])\n",
-       "    ),\n",
-       "    ShapedArray(int16[]),\n",
-       "    ShapedArray(int8[])\n",
-       ")\n",
-       "
\n" - ], - "text/plain": [ - "\u001b[1m(\u001b[0m\n", - " \u001b[1;35mBoard\u001b[0m\u001b[1m(\u001b[0m\n", - " \u001b[33mseed\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfactories_per_team\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mmap\u001b[0m=\u001b[1;35mGameMap\u001b[0m\u001b[1m(\u001b[0m\n", - " \u001b[33mrubble\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m64\u001b[0m,\u001b[1;36m64\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mice\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mbool\u001b[1m[\u001b[0m\u001b[1;36m64\u001b[0m,\u001b[1;36m64\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33more\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mbool\u001b[1m[\u001b[0m\u001b[1;36m64\u001b[0m,\u001b[1;36m64\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33msymmetry\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\n", - " \u001b[1m)\u001b[0m,\n", - " \u001b[33mlichen\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m64\u001b[0m,\u001b[1;36m64\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mlichen_strains\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m64\u001b[0m,\u001b[1;36m64\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33munits_map\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint16\u001b[1m[\u001b[0m\u001b[1;36m64\u001b[0m,\u001b[1;36m64\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfactory_map\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m64\u001b[0m,\u001b[1;36m64\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfactory_occupancy_map\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m64\u001b[0m,\u001b[1;36m64\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfactory_pos\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m22\u001b[0m,\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\n", - " \u001b[1m)\u001b[0m,\n", - " \u001b[1;35mUnit\u001b[0m\u001b[1m(\u001b[0m\n", - " \u001b[33munit_type\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33maction_queue\u001b[0m=\u001b[1;35mActionQueue\u001b[0m\u001b[1m(\u001b[0m\n", - " \u001b[33mdata\u001b[0m=\u001b[1;35mUnitAction\u001b[0m\u001b[1m(\u001b[0m\n", - " \u001b[33maction_type\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m,\u001b[1;36m20\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mdirection\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m,\u001b[1;36m20\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mresource_type\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m,\u001b[1;36m20\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mamount\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint16\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m,\u001b[1;36m20\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mrepeat\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint16\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m,\u001b[1;36m20\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mn\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint16\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m,\u001b[1;36m20\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\n", - " \u001b[1m)\u001b[0m,\n", - " \u001b[33mfront\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mrear\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mcount\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\n", - " \u001b[1m)\u001b[0m,\n", - " \u001b[33mteam_id\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33munit_id\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint16\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mpos\u001b[0m=\u001b[1;35mPosition\u001b[0m\u001b[1m(\u001b[0m\u001b[33mpos\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m,\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mcargo\u001b[0m=\u001b[1;35mUnitCargo\u001b[0m\u001b[1m(\u001b[0m\u001b[33mstock\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m,\u001b[1;36m4\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mpower\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m200\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\n", - " \u001b[1m)\u001b[0m,\n", - " \u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint16\u001b[1m[\u001b[0m\u001b[1;36m2000\u001b[0m,\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint16\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[1;35mFactory\u001b[0m\u001b[1m(\u001b[0m\n", - " \u001b[33mteam_id\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33munit_id\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mpos\u001b[0m=\u001b[1;35mPosition\u001b[0m\u001b[1m(\u001b[0m\u001b[33mpos\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m,\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mpower\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mcargo\u001b[0m=\u001b[1;35mUnitCargo\u001b[0m\u001b[1m(\u001b[0m\u001b[33mstock\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m,\u001b[1;36m4\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\u001b[1m)\u001b[0m\n", - " \u001b[1m)\u001b[0m,\n", - " \u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m22\u001b[0m,\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[1;35mTeam\u001b[0m\u001b[1m(\u001b[0m\n", - " \u001b[33mteam_id\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfaction\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33minit_water\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33minit_metal\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfactories_to_place\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfactory_strains\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mn_factory\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mbid\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\n", - " \u001b[1m)\u001b[0m,\n", - " \u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint16\u001b[1m[\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\n", - "\u001b[1m)\u001b[0m\n" - ] - }, - "metadata": {}, - "output_type": "display_data" + "name": "stdout", + "output_type": "stream", + "text": [ + "[Replay Util] Replaying early steps\n", + "[Replay Util] Replaying early steps - Done\n", + "[Replay Util] Replaying 1/750 steps\n", + "[Replay Util] Replaying 2/750 steps\n", + "[Replay Util] Replaying 3/750 steps\n", + "[Replay Util] Replaying 4/750 steps\n", + "[Replay Util] Replaying 5/750 steps\n", + "[Replay Util] Replaying 6/750 steps\n", + "[Replay Util] Replaying 7/750 steps\n", + "[Replay Util] Replaying 8/750 steps\n", + "[Replay Util] Replaying 9/750 steps\n", + "[Replay Util] Replaying 10/750 steps\n", + "[Replay Util] Replaying 11/750 steps\n", + "[Replay Util] Replaying 12/750 steps\n", + "[Replay Util] Replaying 13/750 steps\n", + "[Replay Util] Replaying 14/750 steps\n", + "[Replay Util] Replaying 15/750 steps\n", + "[Replay Util] Replaying 16/750 steps\n", + "[Replay Util] Replaying 17/750 steps\n", + "[Replay Util] Replaying 18/750 steps\n", + "[Replay Util] Replaying 19/750 steps\n", + "[Replay Util] Replaying 20/750 steps\n", + "[Replay Util] Replaying 21/750 steps\n", + "[Replay Util] Replaying 22/750 steps\n", + "[Replay Util] Replaying 23/750 steps\n", + "[Replay Util] Replaying 24/750 steps\n", + "[Replay Util] Replaying 25/750 steps\n", + "[Replay Util] Replaying 26/750 steps\n", + "[Replay Util] Replaying 27/750 steps\n", + "[Replay Util] Replaying 28/750 steps\n", + "[Replay Util] Replaying 29/750 steps\n", + "[Replay Util] Replaying 30/750 steps\n", + "[Replay Util] Replaying 31/750 steps\n", + "[Replay Util] Replaying 32/750 steps\n", + "[Replay Util] Replaying 33/750 steps\n", + "[Replay Util] Replaying 34/750 steps\n", + "[Replay Util] Replaying 35/750 steps\n", + "[Replay Util] Replaying 36/750 steps\n", + "[Replay Util] Replaying 37/750 steps\n", + "[Replay Util] Replaying 38/750 steps\n", + "[Replay Util] Replaying 39/750 steps\n", + "[Replay Util] Replaying 40/750 steps\n", + "[Replay Util] Replaying 41/750 steps\n", + "[Replay Util] Replaying 42/750 steps\n", + "[Replay Util] Replaying 43/750 steps\n", + "[Replay Util] Replaying 44/750 steps\n", + "[Replay Util] Replaying 45/750 steps\n", + "[Replay Util] Replaying 46/750 steps\n", + "[Replay Util] Replaying 47/750 steps\n", + "[Replay Util] Replaying 48/750 steps\n", + "[Replay Util] Replaying 49/750 steps\n", + "[Replay Util] Replaying 50/750 steps\n", + "[Replay Util] Replaying 51/750 steps\n", + "[Replay Util] Replaying 52/750 steps\n", + "[Replay Util] Replaying 53/750 steps\n", + "[Replay Util] Replaying 54/750 steps\n", + "[Replay Util] Replaying 55/750 steps\n", + "[Replay Util] Replaying 56/750 steps\n", + "[Replay Util] Replaying 57/750 steps\n", + "[Replay Util] Replaying 58/750 steps\n", + "[Replay Util] Replaying 59/750 steps\n", + "[Replay Util] Replaying 60/750 steps\n", + "[Replay Util] Replaying 61/750 steps\n", + "[Replay Util] Replaying 62/750 steps\n", + "[Replay Util] Replaying 63/750 steps\n", + "[Replay Util] Replaying 64/750 steps\n", + "[Replay Util] Replaying 65/750 steps\n", + "[Replay Util] Replaying 66/750 steps\n", + "[Replay Util] Replaying 67/750 steps\n", + "[Replay Util] Replaying 68/750 steps\n", + "[Replay Util] Replaying 69/750 steps\n", + "[Replay Util] Replaying 70/750 steps\n", + "[Replay Util] Replaying 71/750 steps\n", + "[Replay Util] Replaying 72/750 steps\n", + "[Replay Util] Replaying 73/750 steps\n", + "[Replay Util] Replaying 74/750 steps\n", + "[Replay Util] Replaying 75/750 steps\n", + "[Replay Util] Replaying 76/750 steps\n", + "[Replay Util] Replaying 77/750 steps\n", + "[Replay Util] Replaying 78/750 steps\n", + "[Replay Util] Replaying 79/750 steps\n", + "[Replay Util] Replaying 80/750 steps\n", + "[Replay Util] Replaying 81/750 steps\n", + "[Replay Util] Replaying 82/750 steps\n", + "[Replay Util] Replaying 83/750 steps\n", + "[Replay Util] Replaying 84/750 steps\n", + "[Replay Util] Replaying 85/750 steps\n", + "[Replay Util] Replaying 86/750 steps\n", + "[Replay Util] Replaying 87/750 steps\n", + "[Replay Util] Replaying 88/750 steps\n", + "[Replay Util] Replaying 89/750 steps\n", + "[Replay Util] Replaying 90/750 steps\n", + "[Replay Util] Replaying 91/750 steps\n", + "[Replay Util] Replaying 92/750 steps\n", + "[Replay Util] Replaying 93/750 steps\n", + "[Replay Util] Replaying 94/750 steps\n", + "[Replay Util] Replaying 95/750 steps\n", + "[Replay Util] Replaying 96/750 steps\n", + "[Replay Util] Replaying 97/750 steps\n", + "[Replay Util] Replaying 98/750 steps\n", + "[Replay Util] Replaying 99/750 steps\n", + "[Replay Util] Replaying 100/750 steps\n", + "[Replay Util] Replaying 101/750 steps\n", + "[Replay Util] Replaying 102/750 steps\n", + "[Replay Util] Replaying 103/750 steps\n", + "[Replay Util] Replaying 104/750 steps\n", + "[Replay Util] Replaying 105/750 steps\n", + "[Replay Util] Replaying 106/750 steps\n", + "[Replay Util] Replaying 107/750 steps\n", + "[Replay Util] Replaying 108/750 steps\n", + "[Replay Util] Replaying 109/750 steps\n", + "[Replay Util] Replaying 110/750 steps\n", + "[Replay Util] Replaying 111/750 steps\n", + "[Replay Util] Replaying 112/750 steps\n", + "[Replay Util] Replaying 113/750 steps\n", + "[Replay Util] Replaying 114/750 steps\n", + "[Replay Util] Replaying 115/750 steps\n", + "[Replay Util] Replaying 116/750 steps\n", + "[Replay Util] Replaying 117/750 steps\n", + "[Replay Util] Replaying 118/750 steps\n", + "[Replay Util] Replaying 119/750 steps\n", + "[Replay Util] Replaying 120/750 steps\n", + "[Replay Util] Replaying 121/750 steps\n", + "[Replay Util] Replaying 122/750 steps\n", + "[Replay Util] Replaying 123/750 steps\n", + "[Replay Util] Replaying 124/750 steps\n", + "[Replay Util] Replaying 125/750 steps\n", + "[Replay Util] Replaying 126/750 steps\n", + "[Replay Util] Replaying 127/750 steps\n", + "[Replay Util] Replaying 128/750 steps\n", + "[Replay Util] Replaying 129/750 steps\n", + "[Replay Util] Replaying 130/750 steps\n", + "[Replay Util] Replaying 131/750 steps\n", + "[Replay Util] Replaying 132/750 steps\n", + "[Replay Util] Replaying 133/750 steps\n", + "[Replay Util] Replaying 134/750 steps\n", + "[Replay Util] Replaying 135/750 steps\n", + "[Replay Util] Replaying 136/750 steps\n", + "[Replay Util] Replaying 137/750 steps\n", + "[Replay Util] Replaying 138/750 steps\n", + "[Replay Util] Replaying 139/750 steps\n", + "[Replay Util] Replaying 140/750 steps\n", + "[Replay Util] Replaying 141/750 steps\n", + "[Replay Util] Replaying 142/750 steps\n", + "[Replay Util] Replaying 143/750 steps\n", + "[Replay Util] Replaying 144/750 steps\n", + "[Replay Util] Replaying 145/750 steps\n", + "[Replay Util] Replaying 146/750 steps\n", + "[Replay Util] Replaying 147/750 steps\n", + "[Replay Util] Replaying 148/750 steps\n", + "[Replay Util] Replaying 149/750 steps\n", + "[Replay Util] Replaying 150/750 steps\n", + "[Replay Util] Replaying 151/750 steps\n", + "[Replay Util] Replaying 152/750 steps\n", + "[Replay Util] Replaying 153/750 steps\n", + "[Replay Util] Replaying 154/750 steps\n", + "[Replay Util] Replaying 155/750 steps\n", + "[Replay Util] Replaying 156/750 steps\n", + "[Replay Util] Replaying 157/750 steps\n", + "[Replay Util] Replaying 158/750 steps\n", + "[Replay Util] Replaying 159/750 steps\n", + "[Replay Util] Replaying 160/750 steps\n", + "[Replay Util] Replaying 161/750 steps\n", + "[Replay Util] Replaying 162/750 steps\n", + "[Replay Util] Replaying 163/750 steps\n", + "[Replay Util] Replaying 164/750 steps\n", + "[Replay Util] Replaying 165/750 steps\n", + "[Replay Util] Replaying 166/750 steps\n", + "[Replay Util] Replaying 167/750 steps\n", + "[Replay Util] Replaying 168/750 steps\n", + "[Replay Util] Replaying 169/750 steps\n", + "[Replay Util] Replaying 170/750 steps\n", + "[Replay Util] Replaying 171/750 steps\n", + "[Replay Util] Replaying 172/750 steps\n", + "[Replay Util] Replaying 173/750 steps\n", + "[Replay Util] Replaying 174/750 steps\n", + "[Replay Util] Replaying 175/750 steps\n", + "[Replay Util] Replaying 176/750 steps\n", + "[Replay Util] Replaying 177/750 steps\n", + "[Replay Util] Replaying 178/750 steps\n", + "[Replay Util] Replaying 179/750 steps\n", + "[Replay Util] Replaying 180/750 steps\n", + "[Replay Util] Replaying 181/750 steps\n", + "[Replay Util] Replaying 182/750 steps\n", + "[Replay Util] Replaying 183/750 steps\n", + "[Replay Util] Replaying 184/750 steps\n", + "[Replay Util] Replaying 185/750 steps\n", + "[Replay Util] Replaying 186/750 steps\n", + "[Replay Util] Replaying 187/750 steps\n", + "[Replay Util] Replaying 188/750 steps\n", + "[Replay Util] Replaying 189/750 steps\n", + "[Replay Util] Replaying 190/750 steps\n", + "[Replay Util] Replaying 191/750 steps\n", + "[Replay Util] Replaying 192/750 steps\n", + "[Replay Util] Replaying 193/750 steps\n", + "[Replay Util] Replaying 194/750 steps\n", + "[Replay Util] Replaying 195/750 steps\n", + "[Replay Util] Replaying 196/750 steps\n", + "[Replay Util] Replaying 197/750 steps\n", + "[Replay Util] Replaying 198/750 steps\n", + "[Replay Util] Replaying 199/750 steps\n", + "[Replay Util] Replaying 200/750 steps\n", + "[Replay Util] Replaying 201/750 steps\n", + "[Replay Util] Replaying 202/750 steps\n", + "[Replay Util] Replaying 203/750 steps\n", + "[Replay Util] Replaying 204/750 steps\n", + "[Replay Util] Replaying 205/750 steps\n", + "[Replay Util] Replaying 206/750 steps\n", + "[Replay Util] Replaying 207/750 steps\n", + "[Replay Util] Replaying 208/750 steps\n", + "[Replay Util] Replaying 209/750 steps\n", + "[Replay Util] Replaying 210/750 steps\n", + "[Replay Util] Replaying 211/750 steps\n", + "[Replay Util] Replaying 212/750 steps\n", + "[Replay Util] Replaying 213/750 steps\n", + "[Replay Util] Replaying 214/750 steps\n", + "[Replay Util] Replaying 215/750 steps\n", + "[Replay Util] Replaying 216/750 steps\n", + "[Replay Util] Replaying 217/750 steps\n", + "[Replay Util] Replaying 218/750 steps\n", + "[Replay Util] Replaying 219/750 steps\n", + "[Replay Util] Replaying 220/750 steps\n", + "[Replay Util] Replaying 221/750 steps\n", + "[Replay Util] Replaying 222/750 steps\n", + "[Replay Util] Replaying 223/750 steps\n", + "[Replay Util] Replaying 224/750 steps\n", + "[Replay Util] Replaying 225/750 steps\n", + "[Replay Util] Replaying 226/750 steps\n", + "[Replay Util] Replaying 227/750 steps\n", + "[Replay Util] Replaying 228/750 steps\n", + "[Replay Util] Replaying 229/750 steps\n", + "[Replay Util] Replaying 230/750 steps\n", + "[Replay Util] Replaying 231/750 steps\n", + "[Replay Util] Replaying 232/750 steps\n", + "[Replay Util] Replaying 233/750 steps\n", + "[Replay Util] Replaying 234/750 steps\n", + "[Replay Util] Replaying 235/750 steps\n", + "[Replay Util] Replaying 236/750 steps\n", + "[Replay Util] Replaying 237/750 steps\n", + "[Replay Util] Replaying 238/750 steps\n", + "[Replay Util] Replaying 239/750 steps\n", + "[Replay Util] Replaying 240/750 steps\n", + "[Replay Util] Replaying 241/750 steps\n", + "[Replay Util] Replaying 242/750 steps\n", + "[Replay Util] Replaying 243/750 steps\n", + "[Replay Util] Replaying 244/750 steps\n", + "[Replay Util] Replaying 245/750 steps\n", + "[Replay Util] Replaying 246/750 steps\n", + "[Replay Util] Replaying 247/750 steps\n", + "[Replay Util] Replaying 248/750 steps\n", + "[Replay Util] Replaying 249/750 steps\n", + "[Replay Util] Replaying 250/750 steps\n", + "[Replay Util] Replaying 251/750 steps\n", + "[Replay Util] Replaying 252/750 steps\n", + "[Replay Util] Replaying 253/750 steps\n", + "[Replay Util] Replaying 254/750 steps\n", + "[Replay Util] Replaying 255/750 steps\n", + "[Replay Util] Replaying 256/750 steps\n", + "[Replay Util] Replaying 257/750 steps\n", + "[Replay Util] Replaying 258/750 steps\n", + "[Replay Util] Replaying 259/750 steps\n", + "[Replay Util] Replaying 260/750 steps\n", + "[Replay Util] Replaying 261/750 steps\n", + "[Replay Util] Replaying 262/750 steps\n", + "[Replay Util] Replaying 263/750 steps\n", + "[Replay Util] Replaying 264/750 steps\n", + "[Replay Util] Replaying 265/750 steps\n", + "[Replay Util] Replaying 266/750 steps\n", + "[Replay Util] Replaying 267/750 steps\n", + "[Replay Util] Replaying 268/750 steps\n", + "[Replay Util] Replaying 269/750 steps\n", + "[Replay Util] Replaying 270/750 steps\n", + "[Replay Util] Replaying 271/750 steps\n", + "[Replay Util] Replaying 272/750 steps\n", + "[Replay Util] Replaying 273/750 steps\n", + "[Replay Util] Replaying 274/750 steps\n", + "[Replay Util] Replaying 275/750 steps\n", + "[Replay Util] Replaying 276/750 steps\n", + "[Replay Util] Replaying 277/750 steps\n", + "[Replay Util] Replaying 278/750 steps\n", + "[Replay Util] Replaying 279/750 steps\n", + "[Replay Util] Replaying 280/750 steps\n", + "[Replay Util] Replaying 281/750 steps\n", + "[Replay Util] Replaying 282/750 steps\n", + "[Replay Util] Replaying 283/750 steps\n", + "[Replay Util] Replaying 284/750 steps\n", + "[Replay Util] Replaying 285/750 steps\n", + "[Replay Util] Replaying 286/750 steps\n", + "[Replay Util] Replaying 287/750 steps\n", + "[Replay Util] Replaying 288/750 steps\n", + "[Replay Util] Replaying 289/750 steps\n", + "[Replay Util] Replaying 290/750 steps\n", + "[Replay Util] Replaying 291/750 steps\n", + "[Replay Util] Replaying 292/750 steps\n", + "[Replay Util] Replaying 293/750 steps\n", + "[Replay Util] Replaying 294/750 steps\n", + "[Replay Util] Replaying 295/750 steps\n", + "[Replay Util] Replaying 296/750 steps\n", + "[Replay Util] Replaying 297/750 steps\n", + "[Replay Util] Replaying 298/750 steps\n", + "[Replay Util] Replaying 299/750 steps\n", + "[Replay Util] Replaying 300/750 steps\n", + "[Replay Util] Replaying 301/750 steps\n", + "[Replay Util] Replaying 302/750 steps\n", + "[Replay Util] Replaying 303/750 steps\n", + "[Replay Util] Replaying 304/750 steps\n", + "[Replay Util] Replaying 305/750 steps\n", + "[Replay Util] Replaying 306/750 steps\n", + "[Replay Util] Replaying 307/750 steps\n", + "[Replay Util] Replaying 308/750 steps\n", + "[Replay Util] Replaying 309/750 steps\n", + "[Replay Util] Replaying 310/750 steps\n", + "[Replay Util] Replaying 311/750 steps\n", + "[Replay Util] Replaying 312/750 steps\n", + "[Replay Util] Replaying 313/750 steps\n", + "[Replay Util] Replaying 314/750 steps\n", + "[Replay Util] Replaying 315/750 steps\n", + "[Replay Util] Replaying 316/750 steps\n", + "[Replay Util] Replaying 317/750 steps\n", + "[Replay Util] Replaying 318/750 steps\n", + "[Replay Util] Replaying 319/750 steps\n", + "[Replay Util] Replaying 320/750 steps\n", + "[Replay Util] Replaying 321/750 steps\n", + "[Replay Util] Replaying 322/750 steps\n", + "[Replay Util] Replaying 323/750 steps\n", + "[Replay Util] Replaying 324/750 steps\n", + "[Replay Util] Replaying 325/750 steps\n", + "[Replay Util] Replaying 326/750 steps\n", + "[Replay Util] Replaying 327/750 steps\n", + "[Replay Util] Replaying 328/750 steps\n", + "[Replay Util] Replaying 329/750 steps\n", + "[Replay Util] Replaying 330/750 steps\n", + "[Replay Util] Replaying 331/750 steps\n", + "[Replay Util] Replaying 332/750 steps\n", + "[Replay Util] Replaying 333/750 steps\n", + "[Replay Util] Replaying 334/750 steps\n", + "[Replay Util] Replaying 335/750 steps\n", + "[Replay Util] Replaying 336/750 steps\n", + "[Replay Util] Replaying 337/750 steps\n", + "[Replay Util] Replaying 338/750 steps\n", + "[Replay Util] Replaying 339/750 steps\n", + "[Replay Util] Replaying 340/750 steps\n", + "[Replay Util] Replaying 341/750 steps\n", + "[Replay Util] Replaying 342/750 steps\n", + "[Replay Util] Replaying 343/750 steps\n", + "[Replay Util] Replaying 344/750 steps\n", + "[Replay Util] Replaying 345/750 steps\n", + "[Replay Util] Replaying 346/750 steps\n", + "[Replay Util] Replaying 347/750 steps\n", + "[Replay Util] Replaying 348/750 steps\n", + "[Replay Util] Replaying 349/750 steps\n", + "[Replay Util] Replaying 350/750 steps\n", + "[Replay Util] Replaying 351/750 steps\n", + "[Replay Util] Replaying 352/750 steps\n", + "[Replay Util] Replaying 353/750 steps\n", + "[Replay Util] Replaying 354/750 steps\n", + "[Replay Util] Replaying 355/750 steps\n", + "[Replay Util] Replaying 356/750 steps\n", + "[Replay Util] Replaying 357/750 steps\n", + "[Replay Util] Replaying 358/750 steps\n", + "[Replay Util] Replaying 359/750 steps\n", + "[Replay Util] Replaying 360/750 steps\n", + "[Replay Util] Replaying 361/750 steps\n", + "[Replay Util] Replaying 362/750 steps\n", + "[Replay Util] Replaying 363/750 steps\n", + "[Replay Util] Replaying 364/750 steps\n", + "[Replay Util] Replaying 365/750 steps\n", + "[Replay Util] Replaying 366/750 steps\n", + "[Replay Util] Replaying 367/750 steps\n", + "[Replay Util] Replaying 368/750 steps\n", + "[Replay Util] Replaying 369/750 steps\n", + "[Replay Util] Replaying 370/750 steps\n", + "[Replay Util] Replaying 371/750 steps\n", + "[Replay Util] Replaying 372/750 steps\n", + "[Replay Util] Replaying 373/750 steps\n", + "[Replay Util] Replaying 374/750 steps\n", + "[Replay Util] Replaying 375/750 steps\n", + "[Replay Util] Replaying 376/750 steps\n", + "[Replay Util] Replaying 377/750 steps\n", + "[Replay Util] Replaying 378/750 steps\n", + "[Replay Util] Replaying 379/750 steps\n", + "[Replay Util] Replaying 380/750 steps\n", + "[Replay Util] Replaying 381/750 steps\n", + "[Replay Util] Replaying 382/750 steps\n", + "[Replay Util] Replaying 383/750 steps\n", + "[Replay Util] Replaying 384/750 steps\n", + "[Replay Util] Replaying 385/750 steps\n", + "[Replay Util] Replaying 386/750 steps\n", + "[Replay Util] Replaying 387/750 steps\n", + "[Replay Util] Replaying 388/750 steps\n", + "[Replay Util] Replaying 389/750 steps\n", + "[Replay Util] Replaying 390/750 steps\n", + "[Replay Util] Replaying 391/750 steps\n", + "[Replay Util] Replaying 392/750 steps\n", + "[Replay Util] Replaying 393/750 steps\n", + "[Replay Util] Replaying 394/750 steps\n", + "[Replay Util] Replaying 395/750 steps\n", + "[Replay Util] Replaying 396/750 steps\n", + "[Replay Util] Replaying 397/750 steps\n", + "[Replay Util] Replaying 398/750 steps\n", + "[Replay Util] Replaying 399/750 steps\n", + "[Replay Util] Replaying 400/750 steps\n", + "[Replay Util] Replaying 401/750 steps\n", + "[Replay Util] Replaying 402/750 steps\n", + "[Replay Util] Replaying 403/750 steps\n", + "[Replay Util] Replaying 404/750 steps\n", + "[Replay Util] Replaying 405/750 steps\n", + "[Replay Util] Replaying 406/750 steps\n", + "[Replay Util] Replaying 407/750 steps\n", + "[Replay Util] Replaying 408/750 steps\n", + "[Replay Util] Replaying 409/750 steps\n", + "[Replay Util] Replaying 410/750 steps\n", + "[Replay Util] Replaying 411/750 steps\n", + "[Replay Util] Replaying 412/750 steps\n", + "[Replay Util] Replaying 413/750 steps\n", + "[Replay Util] Replaying 414/750 steps\n", + "[Replay Util] Replaying 415/750 steps\n", + "[Replay Util] Replaying 416/750 steps\n", + "[Replay Util] Replaying 417/750 steps\n", + "[Replay Util] Replaying 418/750 steps\n", + "[Replay Util] Replaying 419/750 steps\n", + "[Replay Util] Replaying 420/750 steps\n", + "[Replay Util] Replaying 421/750 steps\n", + "[Replay Util] Replaying 422/750 steps\n", + "[Replay Util] Replaying 423/750 steps\n", + "[Replay Util] Replaying 424/750 steps\n", + "[Replay Util] Replaying 425/750 steps\n", + "[Replay Util] Replaying 426/750 steps\n", + "[Replay Util] Replaying 427/750 steps\n", + "[Replay Util] Replaying 428/750 steps\n", + "[Replay Util] Replaying 429/750 steps\n", + "[Replay Util] Replaying 430/750 steps\n", + "[Replay Util] Replaying 431/750 steps\n", + "[Replay Util] Replaying 432/750 steps\n", + "[Replay Util] Replaying 433/750 steps\n", + "[Replay Util] Replaying 434/750 steps\n", + "[Replay Util] Replaying 435/750 steps\n", + "[Replay Util] Replaying 436/750 steps\n", + "[Replay Util] Replaying 437/750 steps\n", + "[Replay Util] Replaying 438/750 steps\n", + "[Replay Util] Replaying 439/750 steps\n", + "[Replay Util] Replaying 440/750 steps\n", + "[Replay Util] Replaying 441/750 steps\n", + "[Replay Util] Replaying 442/750 steps\n", + "[Replay Util] Replaying 443/750 steps\n", + "[Replay Util] Replaying 444/750 steps\n", + "[Replay Util] Replaying 445/750 steps\n", + "[Replay Util] Replaying 446/750 steps\n", + "[Replay Util] Replaying 447/750 steps\n", + "[Replay Util] Replaying 448/750 steps\n", + "[Replay Util] Replaying 449/750 steps\n", + "[Replay Util] Replaying 450/750 steps\n", + "[Replay Util] Replaying 451/750 steps\n", + "[Replay Util] Replaying 452/750 steps\n", + "[Replay Util] Replaying 453/750 steps\n", + "[Replay Util] Replaying 454/750 steps\n", + "[Replay Util] Replaying 455/750 steps\n", + "[Replay Util] Replaying 456/750 steps\n", + "[Replay Util] Replaying 457/750 steps\n", + "[Replay Util] Replaying 458/750 steps\n", + "[Replay Util] Replaying 459/750 steps\n", + "[Replay Util] Replaying 460/750 steps\n", + "[Replay Util] Replaying 461/750 steps\n", + "[Replay Util] Replaying 462/750 steps\n", + "[Replay Util] Replaying 463/750 steps\n", + "[Replay Util] Replaying 464/750 steps\n", + "[Replay Util] Replaying 465/750 steps\n", + "[Replay Util] Replaying 466/750 steps\n", + "[Replay Util] Replaying 467/750 steps\n", + "[Replay Util] Replaying 468/750 steps\n", + "[Replay Util] Replaying 469/750 steps\n", + "[Replay Util] Replaying 470/750 steps\n", + "[Replay Util] Replaying 471/750 steps\n", + "[Replay Util] Replaying 472/750 steps\n", + "[Replay Util] Replaying 473/750 steps\n", + "[Replay Util] Replaying 474/750 steps\n", + "[Replay Util] Replaying 475/750 steps\n", + "[Replay Util] Replaying 476/750 steps\n", + "[Replay Util] Replaying 477/750 steps\n", + "[Replay Util] Replaying 478/750 steps\n", + "[Replay Util] Replaying 479/750 steps\n", + "[Replay Util] Replaying 480/750 steps\n", + "[Replay Util] Replaying 481/750 steps\n", + "[Replay Util] Replaying 482/750 steps\n", + "[Replay Util] Replaying 483/750 steps\n", + "[Replay Util] Replaying 484/750 steps\n", + "[Replay Util] Replaying 485/750 steps\n", + "[Replay Util] Replaying 486/750 steps\n", + "[Replay Util] Replaying 487/750 steps\n", + "[Replay Util] Replaying 488/750 steps\n", + "[Replay Util] Replaying 489/750 steps\n", + "[Replay Util] Replaying 490/750 steps\n", + "[Replay Util] Replaying 491/750 steps\n", + "[Replay Util] Replaying 492/750 steps\n", + "[Replay Util] Replaying 493/750 steps\n", + "[Replay Util] Replaying 494/750 steps\n", + "[Replay Util] Replaying 495/750 steps\n", + "[Replay Util] Replaying 496/750 steps\n", + "[Replay Util] Replaying 497/750 steps\n", + "[Replay Util] Replaying 498/750 steps\n", + "[Replay Util] Replaying 499/750 steps\n", + "[Replay Util] Replaying 500/750 steps\n", + "[Replay Util] Replaying 501/750 steps\n", + "[Replay Util] Replaying 502/750 steps\n", + "[Replay Util] Replaying 503/750 steps\n", + "[Replay Util] Replaying 504/750 steps\n", + "[Replay Util] Replaying 505/750 steps\n", + "[Replay Util] Replaying 506/750 steps\n", + "[Replay Util] Replaying 507/750 steps\n", + "[Replay Util] Replaying 508/750 steps\n", + "[Replay Util] Replaying 509/750 steps\n", + "[Replay Util] Replaying 510/750 steps\n", + "[Replay Util] Replaying 511/750 steps\n", + "[Replay Util] Replaying 512/750 steps\n", + "[Replay Util] Replaying 513/750 steps\n", + "[Replay Util] Replaying 514/750 steps\n", + "[Replay Util] Replaying 515/750 steps\n", + "[Replay Util] Replaying 516/750 steps\n", + "[Replay Util] Replaying 517/750 steps\n", + "[Replay Util] Replaying 518/750 steps\n", + "[Replay Util] Replaying 519/750 steps\n", + "[Replay Util] Replaying 520/750 steps\n", + "[Replay Util] Replaying 521/750 steps\n", + "[Replay Util] Replaying 522/750 steps\n", + "[Replay Util] Replaying 523/750 steps\n", + "[Replay Util] Replaying 524/750 steps\n", + "[Replay Util] Replaying 525/750 steps\n", + "[Replay Util] Replaying 526/750 steps\n", + "[Replay Util] Replaying 527/750 steps\n", + "[Replay Util] Replaying 528/750 steps\n", + "[Replay Util] Replaying 529/750 steps\n", + "[Replay Util] Replaying 530/750 steps\n", + "[Replay Util] Replaying 531/750 steps\n", + "[Replay Util] Replaying 532/750 steps\n", + "[Replay Util] Replaying 533/750 steps\n", + "[Replay Util] Replaying 534/750 steps\n", + "[Replay Util] Replaying 535/750 steps\n", + "[Replay Util] Replaying 536/750 steps\n", + "[Replay Util] Replaying 537/750 steps\n", + "[Replay Util] Replaying 538/750 steps\n", + "[Replay Util] Replaying 539/750 steps\n", + "[Replay Util] Replaying 540/750 steps\n", + "[Replay Util] Replaying 541/750 steps\n", + "[Replay Util] Replaying 542/750 steps\n", + "[Replay Util] Replaying 543/750 steps\n", + "[Replay Util] Replaying 544/750 steps\n", + "[Replay Util] Replaying 545/750 steps\n", + "[Replay Util] Replaying 546/750 steps\n", + "[Replay Util] Replaying 547/750 steps\n", + "[Replay Util] Replaying 548/750 steps\n", + "[Replay Util] Replaying 549/750 steps\n", + "[Replay Util] Replaying 550/750 steps\n", + "[Replay Util] Replaying 551/750 steps\n", + "[Replay Util] Replaying 552/750 steps\n", + "[Replay Util] Replaying 553/750 steps\n", + "[Replay Util] Replaying 554/750 steps\n", + "[Replay Util] Replaying 555/750 steps\n", + "[Replay Util] Replaying 556/750 steps\n", + "[Replay Util] Replaying 557/750 steps\n", + "[Replay Util] Replaying 558/750 steps\n", + "[Replay Util] Replaying 559/750 steps\n", + "[Replay Util] Replaying 560/750 steps\n", + "[Replay Util] Replaying 561/750 steps\n", + "[Replay Util] Replaying 562/750 steps\n", + "[Replay Util] Replaying 563/750 steps\n", + "[Replay Util] Replaying 564/750 steps\n", + "[Replay Util] Replaying 565/750 steps\n", + "[Replay Util] Replaying 566/750 steps\n", + "[Replay Util] Replaying 567/750 steps\n", + "[Replay Util] Replaying 568/750 steps\n", + "[Replay Util] Replaying 569/750 steps\n", + "[Replay Util] Replaying 570/750 steps\n", + "[Replay Util] Replaying 571/750 steps\n", + "[Replay Util] Replaying 572/750 steps\n", + "[Replay Util] Replaying 573/750 steps\n", + "[Replay Util] Replaying 574/750 steps\n", + "[Replay Util] Replaying 575/750 steps\n", + "[Replay Util] Replaying 576/750 steps\n", + "[Replay Util] Replaying 577/750 steps\n", + "[Replay Util] Replaying 578/750 steps\n", + "[Replay Util] Replaying 579/750 steps\n", + "[Replay Util] Replaying 580/750 steps\n", + "[Replay Util] Replaying 581/750 steps\n", + "[Replay Util] Replaying 582/750 steps\n", + "[Replay Util] Replaying 583/750 steps\n", + "[Replay Util] Replaying 584/750 steps\n", + "[Replay Util] Replaying 585/750 steps\n", + "[Replay Util] Replaying 586/750 steps\n", + "[Replay Util] Replaying 587/750 steps\n", + "[Replay Util] Replaying 588/750 steps\n", + "[Replay Util] Replaying 589/750 steps\n", + "[Replay Util] Replaying 590/750 steps\n", + "[Replay Util] Replaying 591/750 steps\n", + "[Replay Util] Replaying 592/750 steps\n", + "[Replay Util] Replaying 593/750 steps\n", + "[Replay Util] Replaying 594/750 steps\n", + "[Replay Util] Replaying 595/750 steps\n", + "[Replay Util] Replaying 596/750 steps\n", + "[Replay Util] Replaying 597/750 steps\n", + "[Replay Util] Replaying 598/750 steps\n", + "[Replay Util] Replaying 599/750 steps\n", + "[Replay Util] Replaying 600/750 steps\n", + "[Replay Util] Replaying 601/750 steps\n", + "[Replay Util] Replaying 602/750 steps\n", + "[Replay Util] Replaying 603/750 steps\n", + "[Replay Util] Replaying 604/750 steps\n", + "[Replay Util] Replaying 605/750 steps\n", + "[Replay Util] Replaying 606/750 steps\n", + "[Replay Util] Replaying 607/750 steps\n", + "[Replay Util] Replaying 608/750 steps\n", + "[Replay Util] Replaying 609/750 steps\n", + "[Replay Util] Replaying 610/750 steps\n", + "[Replay Util] Replaying 611/750 steps\n", + "[Replay Util] Replaying 612/750 steps\n", + "[Replay Util] Replaying 613/750 steps\n", + "[Replay Util] Replaying 614/750 steps\n", + "[Replay Util] Replaying 615/750 steps\n", + "[Replay Util] Replaying 616/750 steps\n", + "[Replay Util] Replaying 617/750 steps\n", + "[Replay Util] Replaying 618/750 steps\n", + "[Replay Util] Replaying 619/750 steps\n", + "[Replay Util] Replaying 620/750 steps\n", + "[Replay Util] Replaying 621/750 steps\n", + "[Replay Util] Replaying 622/750 steps\n", + "[Replay Util] Replaying 623/750 steps\n", + "[Replay Util] Replaying 624/750 steps\n", + "[Replay Util] Replaying 625/750 steps\n", + "[Replay Util] Replaying 626/750 steps\n", + "[Replay Util] Replaying 627/750 steps\n", + "[Replay Util] Replaying 628/750 steps\n", + "[Replay Util] Replaying 629/750 steps\n", + "[Replay Util] Replaying 630/750 steps\n", + "[Replay Util] Replaying 631/750 steps\n", + "[Replay Util] Replaying 632/750 steps\n", + "[Replay Util] Replaying 633/750 steps\n", + "[Replay Util] Replaying 634/750 steps\n", + "[Replay Util] Replaying 635/750 steps\n", + "[Replay Util] Replaying 636/750 steps\n", + "[Replay Util] Replaying 637/750 steps\n", + "[Replay Util] Replaying 638/750 steps\n", + "[Replay Util] Replaying 639/750 steps\n", + "[Replay Util] Replaying 640/750 steps\n", + "[Replay Util] Replaying 641/750 steps\n", + "[Replay Util] Replaying 642/750 steps\n", + "[Replay Util] Replaying 643/750 steps\n", + "[Replay Util] Replaying 644/750 steps\n", + "[Replay Util] Replaying 645/750 steps\n", + "[Replay Util] Replaying 646/750 steps\n", + "[Replay Util] Replaying 647/750 steps\n", + "[Replay Util] Replaying 648/750 steps\n", + "[Replay Util] Replaying 649/750 steps\n", + "[Replay Util] Replaying 650/750 steps\n", + "[Replay Util] Replaying 651/750 steps\n", + "[Replay Util] Replaying 652/750 steps\n", + "[Replay Util] Replaying 653/750 steps\n", + "[Replay Util] Replaying 654/750 steps\n", + "[Replay Util] Replaying 655/750 steps\n", + "[Replay Util] Replaying 656/750 steps\n", + "[Replay Util] Replaying 657/750 steps\n", + "[Replay Util] Replaying 658/750 steps\n", + "[Replay Util] Replaying 659/750 steps\n", + "[Replay Util] Replaying 660/750 steps\n", + "[Replay Util] Replaying 661/750 steps\n", + "[Replay Util] Replaying 662/750 steps\n", + "[Replay Util] Replaying 663/750 steps\n", + "[Replay Util] Replaying 664/750 steps\n", + "[Replay Util] Replaying 665/750 steps\n", + "[Replay Util] Replaying 666/750 steps\n", + "[Replay Util] Replaying 667/750 steps\n", + "[Replay Util] Replaying 668/750 steps\n", + "[Replay Util] Replaying 669/750 steps\n", + "[Replay Util] Replaying 670/750 steps\n", + "[Replay Util] Replaying 671/750 steps\n", + "[Replay Util] Replaying 672/750 steps\n", + "[Replay Util] Replaying 673/750 steps\n", + "[Replay Util] Replaying 674/750 steps\n", + "[Replay Util] Replaying 675/750 steps\n", + "[Replay Util] Replaying 676/750 steps\n", + "[Replay Util] Replaying 677/750 steps\n", + "[Replay Util] Replaying 678/750 steps\n", + "[Replay Util] Replaying 679/750 steps\n", + "[Replay Util] Replaying 680/750 steps\n", + "[Replay Util] Replaying 681/750 steps\n", + "[Replay Util] Replaying 682/750 steps\n", + "[Replay Util] Replaying 683/750 steps\n", + "[Replay Util] Replaying 684/750 steps\n", + "[Replay Util] Replaying 685/750 steps\n", + "[Replay Util] Replaying 686/750 steps\n", + "[Replay Util] Replaying 687/750 steps\n", + "[Replay Util] Replaying 688/750 steps\n", + "[Replay Util] Replaying 689/750 steps\n", + "[Replay Util] Replaying 690/750 steps\n", + "[Replay Util] Replaying 691/750 steps\n", + "[Replay Util] Replaying 692/750 steps\n", + "[Replay Util] Replaying 693/750 steps\n", + "[Replay Util] Replaying 694/750 steps\n", + "[Replay Util] Replaying 695/750 steps\n", + "[Replay Util] Replaying 696/750 steps\n", + "[Replay Util] Replaying 697/750 steps\n", + "[Replay Util] Replaying 698/750 steps\n", + "[Replay Util] Replaying 699/750 steps\n", + "[Replay Util] Replaying 700/750 steps\n", + "[Replay Util] Replaying 701/750 steps\n", + "[Replay Util] Replaying 702/750 steps\n", + "[Replay Util] Replaying 703/750 steps\n", + "[Replay Util] Replaying 704/750 steps\n", + "[Replay Util] Replaying 705/750 steps\n", + "[Replay Util] Replaying 706/750 steps\n", + "[Replay Util] Replaying 707/750 steps\n", + "[Replay Util] Replaying 708/750 steps\n", + "[Replay Util] Replaying 709/750 steps\n", + "[Replay Util] Replaying 710/750 steps\n", + "[Replay Util] Replaying 711/750 steps\n", + "[Replay Util] Replaying 712/750 steps\n", + "[Replay Util] Replaying 713/750 steps\n", + "[Replay Util] Replaying 714/750 steps\n", + "[Replay Util] Replaying 715/750 steps\n", + "[Replay Util] Replaying 716/750 steps\n", + "[Replay Util] Replaying 717/750 steps\n", + "[Replay Util] Replaying 718/750 steps\n", + "[Replay Util] Replaying 719/750 steps\n", + "[Replay Util] Replaying 720/750 steps\n", + "[Replay Util] Replaying 721/750 steps\n", + "[Replay Util] Replaying 722/750 steps\n", + "[Replay Util] Replaying 723/750 steps\n", + "[Replay Util] Replaying 724/750 steps\n", + "[Replay Util] Replaying 725/750 steps\n", + "[Replay Util] Replaying 726/750 steps\n", + "[Replay Util] Replaying 727/750 steps\n", + "[Replay Util] Replaying 728/750 steps\n", + "[Replay Util] Replaying 729/750 steps\n", + "[Replay Util] Replaying 730/750 steps\n", + "[Replay Util] Replaying 731/750 steps\n", + "[Replay Util] Replaying 732/750 steps\n", + "[Replay Util] Replaying 733/750 steps\n", + "[Replay Util] Replaying 734/750 steps\n", + "[Replay Util] Replaying 735/750 steps\n", + "[Replay Util] Replaying 736/750 steps\n", + "[Replay Util] Replaying 737/750 steps\n", + "[Replay Util] Replaying 738/750 steps\n", + "[Replay Util] Replaying 739/750 steps\n", + "[Replay Util] Replaying 740/750 steps\n", + "[Replay Util] Replaying 741/750 steps\n", + "[Replay Util] Replaying 742/750 steps\n", + "[Replay Util] Replaying 743/750 steps\n", + "[Replay Util] Replaying 744/750 steps\n", + "[Replay Util] Replaying 745/750 steps\n", + "[Replay Util] Replaying 746/750 steps\n", + "[Replay Util] Replaying 747/750 steps\n", + "[Replay Util] Replaying 748/750 steps\n", + "[Replay Util] Replaying 749/750 steps\n", + "[Replay Util] Replaying 750/750 steps\n" + ] } ], "source": [ - "rich.print(jax.tree_map(lambda x: getattr(x, 'aval'), state[4:]))" + "reload(ut)\n", + "lux_env, lux_actions = jux.utils.load_replay('../data/52958192.json')\n", + "jux_env, state = JuxEnv.from_lux(lux_env, buf_cfg=JuxBufferConfig(MAX_N_UNITS=1000))\n", + "\n", + "state, lux_actions = ut.replay_run_early_phase(jux_env, state, lux_actions)\n", + "state, lux_actions = ut.replay_run_n_late_game_step(750, jux_env, state, lux_actions)" ] }, { "cell_type": "code", - "execution_count": 47, + "execution_count": 21, "metadata": {}, "outputs": [], "source": [ @@ -211,54 +832,17 @@ "from functools import partial\n", "\n", "reload(pp)\n", - "unit_existence_map = pp.get_unit_existence(state.unit_mask, state.units.unit_type, state.units.pos.x, state.units.pos.y)" + "unit_feature = pp.get_unit_feature(state)" ] }, { "cell_type": "code", - "execution_count": 48, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "fig, axes = plt.subplots(2, 2, figsize=(8, 8))\n", - "unit_type = [\"light\", \"heavy\"]\n", - "for i in range(2):\n", - " for j in range(2):\n", - " axes[i, j].imshow(unit_existence_map[i, :, :, j], cmap='gray')\n", - " axes[i, j].set_title(f\"Player {i}, Type {unit_type[j]}\")\n", - "fig.suptitle(\"Unit Existence Map\")\n", - "plt.show()" - ] - }, - { - "cell_type": "code", - "execution_count": 72, - "metadata": {}, - "outputs": [], - "source": [ - "reload(pp)\n", - "unit_feature = pp.get_unit_feature(state.unit_mask, state.units.unit_type, state.units.cargo.stock, state.units.power, state.units.pos.x, state.units.pos.y)" - ] - }, - { - "cell_type": "code", - "execution_count": 74, + "execution_count": 22, "metadata": {}, "outputs": [ { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -280,192 +864,66 @@ }, { "cell_type": "code", - "execution_count": 22, + "execution_count": 5, "metadata": {}, "outputs": [ { - "data": { - "text/plain": [ - "ShapedArray(float32[64,64,4])" - ] - }, - "execution_count": 22, - "metadata": {}, - "output_type": "execute_result" + "name": "stdout", + "output_type": "stream", + "text": [ + "[Replay Util] Replaying 1/1 steps\n" + ] } ], "source": [ - "unit_map.aval" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "#### Guarantees\n", - "```python\n", - "observations['player_0'] == observations['player_1'] == state\n", - "dones['player_0'] == dones['player_1']\n", - "infos['player_0'] == infos['player_1'] == {}\n", - "```" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Understanding the `State`\n", - "\n", - "`State` object is a nested `NamedTuple`, with all leaves being `jax.numpy.ndarray`. It has following fields. " + "previous_state = state\n", + "state, lux_actions = ut.replay_run_n_late_game_step(1, jux_env, state, lux_actions)" ] }, { "cell_type": "code", - "execution_count": 14, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "('env_cfg',\n", - " 'seed',\n", - " 'rng_state',\n", - " 'env_steps',\n", - " 'board',\n", - " 'units',\n", - " 'unit_id2idx',\n", - " 'n_units',\n", - " 'factories',\n", - " 'factory_id2idx',\n", - " 'n_factories',\n", - " 'teams',\n", - " 'global_id',\n", - " 'place_first')" - ] - }, - "execution_count": 14, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "state._fields" - ] - }, - { - "cell_type": "markdown", + "execution_count": 10, "metadata": {}, + "outputs": [], "source": [ - "### Board Information\n", - "\n", - "Information about the board, including rubble, ice, ore, and lichen, are stored in `state.board`." + "reload(pp)\n", + "factory_feature = pp.get_factory_feature(state, previous_state.factories.power)" ] }, { "cell_type": "code", - "execution_count": 13, + "execution_count": 11, "metadata": {}, "outputs": [ { "data": { - "text/html": [ - "
Board(\n",
-       "    seed=ShapedArray(int32[]),\n",
-       "    factories_per_team=ShapedArray(int8[]),\n",
-       "    map=GameMap(\n",
-       "        rubble=ShapedArray(int8[48,48]),\n",
-       "        ice=ShapedArray(bool[48,48]),\n",
-       "        ore=ShapedArray(bool[48,48]),\n",
-       "        symmetry=ShapedArray(int8[])\n",
-       "    ),\n",
-       "    lichen=ShapedArray(int32[48,48]),\n",
-       "    lichen_strains=ShapedArray(int8[48,48]),\n",
-       "    units_map=ShapedArray(int16[48,48]),\n",
-       "    factory_map=ShapedArray(int8[48,48]),\n",
-       "    factory_occupancy_map=ShapedArray(int8[48,48]),\n",
-       "    factory_pos=ShapedArray(int8[22,2])\n",
-       ")\n",
-       "
\n" - ], "text/plain": [ - "\u001b[1;35mBoard\u001b[0m\u001b[1m(\u001b[0m\n", - " \u001b[33mseed\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfactories_per_team\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mmap\u001b[0m=\u001b[1;35mGameMap\u001b[0m\u001b[1m(\u001b[0m\n", - " \u001b[33mrubble\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m48\u001b[0m,\u001b[1;36m48\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mice\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mbool\u001b[1m[\u001b[0m\u001b[1;36m48\u001b[0m,\u001b[1;36m48\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33more\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mbool\u001b[1m[\u001b[0m\u001b[1;36m48\u001b[0m,\u001b[1;36m48\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33msymmetry\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\n", - " \u001b[1m)\u001b[0m,\n", - " \u001b[33mlichen\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m48\u001b[0m,\u001b[1;36m48\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mlichen_strains\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m48\u001b[0m,\u001b[1;36m48\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33munits_map\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint16\u001b[1m[\u001b[0m\u001b[1;36m48\u001b[0m,\u001b[1;36m48\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfactory_map\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m48\u001b[0m,\u001b[1;36m48\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfactory_occupancy_map\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m48\u001b[0m,\u001b[1;36m48\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mfactory_pos\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m22\u001b[0m,\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\n", - "\u001b[1m)\u001b[0m\n" + "(2, 64, 64, 7)" ] }, + "execution_count": 11, "metadata": {}, - "output_type": "display_data" + "output_type": "execute_result" } ], "source": [ - "import jux.tree_util\n", - "import rich\n", - "rich.print(jux.tree_util.map_to_aval(state.board))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Factory Information\n", - "\n", - "All information about factories, including their position, cargo, and power, are stored in `state.factories`. `state.n_factories` indicates the number of factories each player has. Because we have just reset the environment, both players have 0 factory. The leaves of `state.factories` have shapes shown as below. " + "factory_feature.shape" ] }, { "cell_type": "code", - "execution_count": 14, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "state.n_factories = Array([2, 2], dtype=int8)\n" - ] - }, - { - "data": { - "text/html": [ - "
Factory(\n",
-       "    team_id=ShapedArray(int8[2,11]),\n",
-       "    unit_id=ShapedArray(int8[2,11]),\n",
-       "    pos=Position(pos=ShapedArray(int8[2,11,2])),\n",
-       "    power=ShapedArray(int32[2,11]),\n",
-       "    cargo=UnitCargo(stock=ShapedArray(int32[2,11,4]))\n",
-       ")\n",
-       "
\n" - ], - "text/plain": [ - "\u001b[1;35mFactory\u001b[0m\u001b[1m(\u001b[0m\n", - " \u001b[33mteam_id\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33munit_id\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mpos\u001b[0m=\u001b[1;35mPosition\u001b[0m\u001b[1m(\u001b[0m\u001b[33mpos\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint8\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m,\u001b[1;36m2\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mpower\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m,\n", - " \u001b[33mcargo\u001b[0m=\u001b[1;35mUnitCargo\u001b[0m\u001b[1m(\u001b[0m\u001b[33mstock\u001b[0m=\u001b[1;35mShapedArray\u001b[0m\u001b[1m(\u001b[0mint32\u001b[1m[\u001b[0m\u001b[1;36m2\u001b[0m,\u001b[1;36m11\u001b[0m,\u001b[1;36m4\u001b[0m\u001b[1m]\u001b[0m\u001b[1m)\u001b[0m\u001b[1m)\u001b[0m\n", - "\u001b[1m)\u001b[0m\n" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], + "outputs": [], "source": [ - "print(f\"{state.n_factories = }\")\n", - "rich.print(jux.tree_util.map_to_aval(state.factories))" + "fig, axes = plt.subplots(2, 7, figsize=(28, 8))\n", + "features = ['existence', 'ice', 'ore', 'water', 'metal', 'power', 'previous_power']\n", + "for i in range(2):\n", + " for j in range(7):\n", + " axes[i, j].imshow(factory_feature[i, :, :, j], cmap='gray')\n", + " axes[i, j].set_title(f\"Player {i}, {features[j]}\")\n", + "fig.suptitle(\"Factory Features\")\n", + "plt.show()\n" ] } ], diff --git a/state.md b/state.md new file mode 100644 index 0000000..c373c53 --- /dev/null +++ b/state.md @@ -0,0 +1,65 @@ + +```python +( + Board( + seed=ShapedArray(int32[]), + factories_per_team=ShapedArray(int8[]), + map=GameMap( + rubble=ShapedArray(int8[64,64]), + ice=ShapedArray(bool[64,64]), + ore=ShapedArray(bool[64,64]), + symmetry=ShapedArray(int8[]) + ), + lichen=ShapedArray(int32[64,64]), + lichen_strains=ShapedArray(int8[64,64]), + units_map=ShapedArray(int16[64,64]), + factory_map=ShapedArray(int8[64,64]), + factory_occupancy_map=ShapedArray(int8[64,64]), + factory_pos=ShapedArray(int8[22,2]) + ), + Unit( + unit_type=ShapedArray(int8[2,200]), + action_queue=ActionQueue( + data=UnitAction( + action_type=ShapedArray(int8[2,200,20]), + direction=ShapedArray(int8[2,200,20]), + resource_type=ShapedArray(int8[2,200,20]), + amount=ShapedArray(int16[2,200,20]), + repeat=ShapedArray(int16[2,200,20]), + n=ShapedArray(int16[2,200,20]) + ), + front=ShapedArray(int8[2,200]), + rear=ShapedArray(int8[2,200]), + count=ShapedArray(int8[2,200]) + ), + team_id=ShapedArray(int8[2,200]), + unit_id=ShapedArray(int16[2,200]), + pos=Position(pos=ShapedArray(int8[2,200,2])), + cargo=UnitCargo(stock=ShapedArray(int32[2,200,4])), + power=ShapedArray(int32[2,200]) + ), + ShapedArray(int16[2000,2]), unit_id2idx + ShapedArray(int16[2]), n_units + Factory( + team_id=ShapedArray(int8[2,11]), + unit_id=ShapedArray(int8[2,11]), + pos=Position(pos=ShapedArray(int8[2,11,2])), + power=ShapedArray(int32[2,11]), + cargo=UnitCargo(stock=ShapedArray(int32[2,11,4])) + ), + ShapedArray(int8[22,2]), factory_id2idx + ShapedArray(int8[2]), n_factories + Team( + team_id=ShapedArray(int8[2]), + faction=ShapedArray(int8[2]), + init_water=ShapedArray(int32[2]), + init_metal=ShapedArray(int32[2]), + factories_to_place=ShapedArray(int32[2]), + factory_strains=ShapedArray(int8[2,11]), + n_factory=ShapedArray(int8[2]), + bid=ShapedArray(int32[2]) + ), + ShapedArray(int16[]), global_id + ShapedArray(int8[]), place_first +) +```