diff --git a/training/eval_history.json b/training/eval_history.json index 2892b9d4..43fc7875 100644 --- a/training/eval_history.json +++ b/training/eval_history.json @@ -522,5 +522,32 @@ "team_1": 44 }, "win_rate_a": 0.47 + }, + { + "timestamp": "2026-08-13T00:34:15+00:00", + "model_a": "/home/jcreek/ai-training/CosmicClash/Game/bots/20260812-1922-gen5-s4-handling-retry1.json", + "model_b": "/home/jcreek/ai-training/CosmicClash/Game/bots/promoted/easy.json", + "seed": 1, + "episodes": 100, + "wins_a": 45, + "wins_b": 30, + "draws": 25, + "side_results": { + "a_team_0": { + "wins_a": 22, + "wins_b": 16, + "draws": 12 + }, + "a_team_1": { + "wins_a": 23, + "wins_b": 14, + "draws": 13 + } + }, + "physical_team_wins": { + "team_0": 36, + "team_1": 39 + }, + "win_rate_a": 0.45 } ] diff --git a/training/generation5_state.json b/training/generation5_state.json index 8eb8ec01..8563feba 100644 --- a/training/generation5_state.json +++ b/training/generation5_state.json @@ -1,6 +1,6 @@ { "stage_index": 0, - "attempt": 1, + "attempt": 2, "status": "in_progress", "log": [ { @@ -87,6 +87,91 @@ } ], "decision": "fail" + }, + { + "stage_index": 0, + "stage_number": 4, + "stage_name": "handling", + "experiment": "20260812-1922-gen5-s4-handling-retry1", + "attempt": 1, + "telemetry_tail": { + "rollout/air_touch_fraction": 0.00025999999418854715, + "rollout/airborne_fraction": 0.43636737924814223, + "rollout/ep_len_mean": 141.96904039001464, + "rollout/ep_rew_mean": -2.6134134426116944, + "rollout/forward_motion_fraction": 0.22919738078117372, + "rollout/goal_rate": 0.5506400021910668, + "rollout/mean_altitude": 4.357596744537354, + "rollout/productive_air_touch_fraction": 7.999999821186065e-05, + "rollout/upright_fraction": 0.2389067138582468, + "rollout/vertical_thrust_mean": 0.05124999895330984 + }, + "telemetry_failures": [ + "rollout/goal_rate=0.5506 < 0.8000", + "rollout/upright_fraction=0.2389 < 0.4500", + "rollout/forward_motion_fraction=0.2292 < 0.2500" + ], + "evaluation_goal_failures": [ + "easy.json: goal_rate=0.750 < 0.800" + ], + "side_balance_failures": [], + "eval": { + "timestamp": "2026-08-13T00:34:15+00:00", + "model_a": "/home/jcreek/ai-training/CosmicClash/Game/bots/20260812-1922-gen5-s4-handling-retry1.json", + "model_b": "/home/jcreek/ai-training/CosmicClash/Game/bots/promoted/easy.json", + "seed": 1, + "episodes": 100, + "wins_a": 45, + "wins_b": 30, + "draws": 25, + "side_results": { + "a_team_0": { + "wins_a": 22, + "wins_b": 16, + "draws": 12 + }, + "a_team_1": { + "wins_a": 23, + "wins_b": 14, + "draws": 13 + } + }, + "physical_team_wins": { + "team_0": 36, + "team_1": 39 + }, + "win_rate_a": 0.45 + }, + "evals": [ + { + "timestamp": "2026-08-13T00:34:15+00:00", + "model_a": "/home/jcreek/ai-training/CosmicClash/Game/bots/20260812-1922-gen5-s4-handling-retry1.json", + "model_b": "/home/jcreek/ai-training/CosmicClash/Game/bots/promoted/easy.json", + "seed": 1, + "episodes": 100, + "wins_a": 45, + "wins_b": 30, + "draws": 25, + "side_results": { + "a_team_0": { + "wins_a": 22, + "wins_b": 16, + "draws": 12 + }, + "a_team_1": { + "wins_a": 23, + "wins_b": 14, + "draws": 13 + } + }, + "physical_team_wins": { + "team_0": 36, + "team_1": 39 + }, + "win_rate_a": 0.45 + } + ], + "decision": "fail" } ] }