Run v2

training/runs/v2/config.json

@@ -0,0 +1,14 @@
+{
+  "W_PER_SHEEP": 2.0,
+  "W_ALIGN": 0.05,
+  "W_PEN_BONUS": 10.0,
+  "W_COMPLETE": 100.0,
+  "W_STEP_COST": 0.02,
+  "W_COMPACT": 0.0,
+  "W_WALL_TOUCH": 0.0,
+  "WALL_TOUCH_BUFFER": 0.4,
+  "ALIGN_SHAPE": "standoff",
+  "ALIGN_GATED": true,
+  "ENTRY_AWARE": true,
+  "ent_coef": 0.02
+}

training/runs/v2/stage_results.json

@@ -0,0 +1,197 @@
+[
+  {
+    "sr": 1.0,
+    "mean_len": 234.0,
+    "mean_min_pen": 3.6668872674306234,
+    "mean_act": 0.4068990752695293,
+    "failure_modes": {
+      "SUCCESS": 30
+    },
+    "reward_per_step": {
+      "progress": 0.11183513424165568,
+      "alignment": 0.0002786317654047819,
+      "compact": 0.0,
+      "wall_touch": 0.0,
+      "pen_bonus": 0.042735042735042736,
+      "step_cost": -0.019999999999999716,
+      "complete": 0.42735042735042733
+    },
+    "n_sheep": 1
+  },
+  {
+    "sr": 0.8666666666666667,
+    "mean_len": 1063.6666666666667,
+    "mean_min_pen": 4.120940693219503,
+    "mean_act": 0.5870139278816712,
+    "failure_modes": {
+      "SUCCESS": 26,
+      "COMPACT_CANT_DRIVE": 4
+    },
+    "reward_per_step": {
+      "progress": 0.05651345582855781,
+      "alignment": 0.007121706701510673,
+      "compact": 0.0,
+      "wall_touch": 0.0,
+      "pen_bonus": 0.01629583202757756,
+      "step_cost": -0.0199999999999909,
+      "complete": 0.08147916013788781
+    },
+    "n_sheep": 2
+  },
+  {
+    "sr": 1.0,
+    "mean_len": 768.6,
+    "mean_min_pen": 3.4802104949951174,
+    "mean_act": 0.7173416881465967,
+    "failure_modes": {
+      "SUCCESS": 30
+    },
+    "reward_per_step": {
+      "progress": 0.11210350058336283,
+      "alignment": 0.007752684222105381,
+      "compact": 0.0,
+      "wall_touch": 0.0,
+      "pen_bonus": 0.039032006245121,
+      "step_cost": -0.019999999999994387,
+      "complete": 0.13010668748373666
+    },
+    "n_sheep": 3
+  },
+  {
+    "sr": 1.0,
+    "mean_len": 749.8666666666667,
+    "mean_min_pen": 3.491257842381795,
+    "mean_act": 1.2302732761302806,
+    "failure_modes": {
+      "SUCCESS": 30
+    },
+    "reward_per_step": {
+      "progress": 0.15859288932254823,
+      "alignment": 0.011327628562653137,
+      "compact": 0.0,
+      "wall_touch": 0.0,
+      "pen_bonus": 0.05334281650071124,
+      "step_cost": -0.0199999999999947,
+      "complete": 0.13335704125177808
+    },
+    "n_sheep": 4
+  },
+  {
+    "sr": 0.9666666666666667,
+    "mean_len": 920.5666666666667,
+    "mean_min_pen": 3.2368871172269187,
+    "mean_act": 1.329068384219205,
+    "failure_modes": {
+      "SUCCESS": 29,
+      "PARTIAL_3of5": 1
+    },
+    "reward_per_step": {
+      "progress": 0.15654392868672135,
+      "alignment": 0.013497823599666012,
+      "compact": 0.0,
+      "wall_touch": 0.0,
+      "pen_bonus": 0.05359017996161784,
+      "step_cost": -0.019999999999992312,
+      "complete": 0.10500778505992686
+    },
+    "n_sheep": 5
+  },
+  {
+    "sr": 0.9666666666666667,
+    "mean_len": 1193.2333333333333,
+    "mean_min_pen": 3.4217512369155885,
+    "mean_act": 1.3575613093489967,
+    "failure_modes": {
+      "COMPACT_CANT_DRIVE": 1,
+      "SUCCESS": 29
+    },
+    "reward_per_step": {
+      "progress": 0.15969395095863717,
+      "alignment": 0.017340700156353795,
+      "compact": 0.0,
+      "wall_touch": 0.0,
+      "pen_bonus": 0.049166131240048046,
+      "step_cost": -0.01999999999998991,
+      "complete": 0.08101237533871554
+    },
+    "n_sheep": 6
+  },
+  {
+    "sr": 1.0,
+    "mean_len": 1209.4666666666667,
+    "mean_min_pen": 3.2339003403981526,
+    "mean_act": 1.3714931576761524,
+    "failure_modes": {
+      "SUCCESS": 30
+    },
+    "reward_per_step": {
+      "progress": 0.17738547200352864,
+      "alignment": 0.017914342656107935,
+      "compact": 0.0,
+      "wall_touch": 0.0,
+      "pen_bonus": 0.057876750082681075,
+      "step_cost": -0.019999999999989804,
+      "complete": 0.08268107154668725
+    },
+    "n_sheep": 7
+  },
+  {
+    "sr": 1.0,
+    "mean_len": 1491.7666666666667,
+    "mean_min_pen": 3.216744065284729,
+    "mean_act": 1.3783802580111435,
+    "failure_modes": {
+      "SUCCESS": 30
+    },
+    "reward_per_step": {
+      "progress": 0.19162546125035912,
+      "alignment": 0.018971863842493202,
+      "compact": 0.0,
+      "wall_touch": 0.0,
+      "pen_bonus": 0.05362768976381472,
+      "step_cost": -0.01999999999998829,
+      "complete": 0.06703461220476839
+    },
+    "n_sheep": 8
+  },
+  {
+    "sr": 0.9,
+    "mean_len": 1627.5666666666666,
+    "mean_min_pen": 3.23857311407725,
+    "mean_act": 1.3832202011732966,
+    "failure_modes": {
+      "SUCCESS": 27,
+      "COMPACT_CANT_DRIVE": 3
+    },
+    "reward_per_step": {
+      "progress": 0.18015228593205654,
+      "alignment": 0.020407598899987247,
+      "compact": 0.0,
+      "wall_touch": 0.0,
+      "pen_bonus": 0.05140598439388044,
+      "step_cost": -0.01999999999998775,
+      "complete": 0.055297274049194094
+    },
+    "n_sheep": 9
+  },
+  {
+    "sr": 0.9333333333333333,
+    "mean_len": 1869.9666666666667,
+    "mean_min_pen": 3.1344878753026326,
+    "mean_act": 1.3841143385300063,
+    "failure_modes": {
+      "SUCCESS": 28,
+      "COMPACT_CANT_DRIVE": 2
+    },
+    "reward_per_step": {
+      "progress": 0.17267533684098152,
+      "alignment": 0.021850885374692264,
+      "compact": 0.0,
+      "wall_touch": 0.0,
+      "pen_bonus": 0.05222909499278062,
+      "step_cost": -0.019999999999986983,
+      "complete": 0.04991176313303267
+    },
+    "n_sheep": 10
+  }
+]