paulh
/
donkeycar-rl-autoresearch
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

wave3: autoresearch trial 15 results 

Agent: pi
Tests: N/A
Tests-Added: 0
TypeScript: N/A
This commit is contained in:
Paul Huliganga 2026-04-16 08:43:17 -04:00
parent 0505de7e63
commit 45b057e9c1
2 changed files with 116 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

111
agent/outerloop-results/autoresearch_phase4_log.txt
View File

@ -488,3 +488,114 @@
[2026-04-16 03:31:41] score=176.67 params={'learning_rate': 0.0016223486895735558, 'steps_per_switch': 5524, 'total_timesteps': 79532} [2026-04-16 03:31:41] score=176.67 params={'learning_rate': 0.0016223486895735558, 'steps_per_switch': 5524, 'total_timesteps': 79532}
[2026-04-16 03:31:41] score=141.35 params={'learning_rate': 0.0010578566658808422, 'steps_per_switch': 4587, 'total_timesteps': 89501} [2026-04-16 03:31:41] score=141.35 params={'learning_rate': 0.0010578566658808422, 'steps_per_switch': 4587, 'total_timesteps': 89501}
[2026-04-16 03:31:41] score=116.26 params={'learning_rate': 0.0008478421211922181, 'steps_per_switch': 14326, 'total_timesteps': 89798} [2026-04-16 03:31:41] score=116.26 params={'learning_rate': 0.0008478421211922181, 'steps_per_switch': 14326, 'total_timesteps': 89798}
[2026-04-16 03:31:42] [Wave4] ✅ Git push complete after trial 10
[2026-04-16 03:31:44]
[Wave4] ========== Trial 11/25 ==========
[2026-04-16 03:31:44] [Wave4] GP UCB top-5 proposals:
[2026-04-16 03:31:44] UCB=2.3190 mu=1.6261 σ=0.3465 params={'learning_rate': 0.0004453206607211525, 'steps_per_switch': 6345, 'total_timesteps': 89491}
[2026-04-16 03:31:44] UCB=2.1595 mu=1.1793 σ=0.4901 params={'learning_rate': 0.00046964743340115834, 'steps_per_switch': 4776, 'total_timesteps': 89856}
[2026-04-16 03:31:44] UCB=2.0941 mu=0.7922 σ=0.6510 params={'learning_rate': 0.0003410660041988607, 'steps_per_switch': 4166, 'total_timesteps': 89400}
[2026-04-16 03:31:44] UCB=2.0525 mu=0.6500 σ=0.7012 params={'learning_rate': 0.00010060597179209699, 'steps_per_switch': 6926, 'total_timesteps': 89756}
[2026-04-16 03:31:44] UCB=2.0368 mu=0.7629 σ=0.6370 params={'learning_rate': 0.00012830012767488515, 'steps_per_switch': 6077, 'total_timesteps': 88482}
[2026-04-16 03:31:44] [Wave4] Proposed params: {'learning_rate': 0.0004453206607211525, 'steps_per_switch': 6345, 'total_timesteps': 89491}
[2026-04-16 03:31:46] [Wave4] Launching trial 11: {'learning_rate': 0.0004453206607211525, 'steps_per_switch': 6345, 'total_timesteps': 89491}
[2026-04-16 03:31:46] [Wave4] Command: python3 /home/paulh/projects/donkeycar-rl-autoresearch/agent/multitrack_runner.py --total-timesteps 89491 --steps-per-switch 6345 --learning-rate 0.0004453206607211525 --eval-episodes 3 --save-dir /home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0011
[2026-04-16 04:47:26] [Wave4] Trial 11 finished in 4539.4s, rc=0
[2026-04-16 04:47:26] [Wave4] Parsed: combined=85.169 mini_monaco=85.169
[2026-04-16 04:47:26] [Wave4] ===== Trial 11 Summary =====
[2026-04-16 04:47:26] GP data points : 11
[2026-04-16 04:47:26] Wave4 Champion: trial=3 score=1943.10 params={'learning_rate': 0.0006852550685205609, 'steps_per_switch': 17499, 'total_timesteps': 157743}
[2026-04-16 04:47:26] Top 5:
[2026-04-16 04:47:26] score=1943.10 params={'learning_rate': 0.0006852550685205609, 'steps_per_switch': 17499, 'total_timesteps': 90000}
[2026-04-16 04:47:26] score=1435.04 params={'learning_rate': 0.0007252855740444645, 'steps_per_switch': 6851, 'total_timesteps': 89893}
[2026-04-16 04:47:26] score=176.67 params={'learning_rate': 0.0016223486895735558, 'steps_per_switch': 5524, 'total_timesteps': 79532}
[2026-04-16 04:47:26] score=141.35 params={'learning_rate': 0.0010578566658808422, 'steps_per_switch': 4587, 'total_timesteps': 89501}
[2026-04-16 04:47:26] score=116.26 params={'learning_rate': 0.0008478421211922181, 'steps_per_switch': 14326, 'total_timesteps': 89798}
[2026-04-16 04:47:28]
[Wave4] ========== Trial 12/25 ==========
[2026-04-16 04:47:28] [Wave4] GP UCB top-5 proposals:
[2026-04-16 04:47:28] UCB=2.2712 mu=2.0464 σ=0.1124 params={'learning_rate': 0.0008597472207162354, 'steps_per_switch': 6936, 'total_timesteps': 89806}
[2026-04-16 04:47:28] UCB=2.1972 mu=0.3727 σ=0.9123 params={'learning_rate': 0.00011512144682169134, 'steps_per_switch': 7128, 'total_timesteps': 67694}
[2026-04-16 04:47:28] UCB=2.1921 mu=0.3206 σ=0.9357 params={'learning_rate': 0.00011280326159121539, 'steps_per_switch': 3550, 'total_timesteps': 67913}
[2026-04-16 04:47:28] UCB=2.1765 mu=1.4350 σ=0.3708 params={'learning_rate': 0.00048039814596614487, 'steps_per_switch': 14975, 'total_timesteps': 84957}
[2026-04-16 04:47:28] UCB=2.1664 mu=0.3542 σ=0.9061 params={'learning_rate': 0.00022405059432424194, 'steps_per_switch': 5228, 'total_timesteps': 67529}
[2026-04-16 04:47:28] [Wave4] Proposed params: {'learning_rate': 0.0008597472207162354, 'steps_per_switch': 6936, 'total_timesteps': 89806}
[2026-04-16 04:47:30] [Wave4] Launching trial 12: {'learning_rate': 0.0008597472207162354, 'steps_per_switch': 6936, 'total_timesteps': 89806}
[2026-04-16 04:47:30] [Wave4] Command: python3 /home/paulh/projects/donkeycar-rl-autoresearch/agent/multitrack_runner.py --total-timesteps 89806 --steps-per-switch 6936 --learning-rate 0.0008597472207162354 --eval-episodes 3 --save-dir /home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0012
[2026-04-16 06:01:18] [Wave4] Trial 12 finished in 4428.4s, rc=0
[2026-04-16 06:01:18] [Wave4] Parsed: combined=131.9914 mini_monaco=131.9914
[2026-04-16 06:01:18] [Wave4] ===== Trial 12 Summary =====
[2026-04-16 06:01:18] GP data points : 12
[2026-04-16 06:01:18] Wave4 Champion: trial=3 score=1943.10 params={'learning_rate': 0.0006852550685205609, 'steps_per_switch': 17499, 'total_timesteps': 157743}
[2026-04-16 06:01:18] Top 5:
[2026-04-16 06:01:18] score=1943.10 params={'learning_rate': 0.0006852550685205609, 'steps_per_switch': 17499, 'total_timesteps': 90000}
[2026-04-16 06:01:18] score=1435.04 params={'learning_rate': 0.0007252855740444645, 'steps_per_switch': 6851, 'total_timesteps': 89893}
[2026-04-16 06:01:18] score=176.67 params={'learning_rate': 0.0016223486895735558, 'steps_per_switch': 5524, 'total_timesteps': 79532}
[2026-04-16 06:01:18] score=141.35 params={'learning_rate': 0.0010578566658808422, 'steps_per_switch': 4587, 'total_timesteps': 89501}
[2026-04-16 06:01:18] score=131.99 params={'learning_rate': 0.0008597472207162354, 'steps_per_switch': 6936, 'total_timesteps': 89806}
[2026-04-16 06:01:20]
[Wave4] ========== Trial 13/25 ==========
[2026-04-16 06:01:20] [Wave4] GP UCB top-5 proposals:
[2026-04-16 06:01:20] UCB=3.6412 mu=2.1560 σ=0.7426 params={'learning_rate': 0.0019124682919611855, 'steps_per_switch': 3574, 'total_timesteps': 73915}
[2026-04-16 06:01:20] UCB=3.5658 mu=2.1987 σ=0.6835 params={'learning_rate': 0.0018623702999958462, 'steps_per_switch': 4072, 'total_timesteps': 73804}
[2026-04-16 06:01:20] UCB=3.5079 mu=2.0033 σ=0.7523 params={'learning_rate': 0.001647357412665117, 'steps_per_switch': 3844, 'total_timesteps': 72144}
[2026-04-16 06:01:20] UCB=3.4793 mu=2.2411 σ=0.6191 params={'learning_rate': 0.0017259517065612783, 'steps_per_switch': 3332, 'total_timesteps': 76276}
[2026-04-16 06:01:20] UCB=3.4632 mu=1.7809 σ=0.8411 params={'learning_rate': 0.0018817829021946482, 'steps_per_switch': 4363, 'total_timesteps': 70258}
[2026-04-16 06:01:20] [Wave4] Proposed params: {'learning_rate': 0.0019124682919611855, 'steps_per_switch': 3574, 'total_timesteps': 73915}
[2026-04-16 06:01:22] [Wave4] Launching trial 13: {'learning_rate': 0.0019124682919611855, 'steps_per_switch': 3574, 'total_timesteps': 73915}
[2026-04-16 06:01:22] [Wave4] Command: python3 /home/paulh/projects/donkeycar-rl-autoresearch/agent/multitrack_runner.py --total-timesteps 73915 --steps-per-switch 3574 --learning-rate 0.0019124682919611855 --eval-episodes 3 --save-dir /home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0013
[2026-04-16 07:01:28] [Wave4] Trial 13 finished in 3605.7s, rc=0
[2026-04-16 07:01:28] [Wave4] Parsed: combined=86.8189 mini_monaco=86.8189
[2026-04-16 07:01:28] [Wave4] ===== Trial 13 Summary =====
[2026-04-16 07:01:28] GP data points : 13
[2026-04-16 07:01:28] Wave4 Champion: trial=3 score=1943.10 params={'learning_rate': 0.0006852550685205609, 'steps_per_switch': 17499, 'total_timesteps': 157743}
[2026-04-16 07:01:28] Top 5:
[2026-04-16 07:01:28] score=1943.10 params={'learning_rate': 0.0006852550685205609, 'steps_per_switch': 17499, 'total_timesteps': 90000}
[2026-04-16 07:01:28] score=1435.04 params={'learning_rate': 0.0007252855740444645, 'steps_per_switch': 6851, 'total_timesteps': 89893}
[2026-04-16 07:01:28] score=176.67 params={'learning_rate': 0.0016223486895735558, 'steps_per_switch': 5524, 'total_timesteps': 79532}
[2026-04-16 07:01:28] score=141.35 params={'learning_rate': 0.0010578566658808422, 'steps_per_switch': 4587, 'total_timesteps': 89501}
[2026-04-16 07:01:28] score=131.99 params={'learning_rate': 0.0008597472207162354, 'steps_per_switch': 6936, 'total_timesteps': 89806}
[2026-04-16 07:01:30]
[Wave4] ========== Trial 14/25 ==========
[2026-04-16 07:01:30] [Wave4] GP UCB top-5 proposals:
[2026-04-16 07:01:30] UCB=3.4396 mu=1.7713 σ=0.8341 params={'learning_rate': 0.00033896311195395496, 'steps_per_switch': 5448, 'total_timesteps': 69359}
[2026-04-16 07:01:30] UCB=3.4124 mu=1.7531 σ=0.8296 params={'learning_rate': 0.00020849841344963834, 'steps_per_switch': 6509, 'total_timesteps': 69502}
[2026-04-16 07:01:30] UCB=3.3878 mu=1.7622 σ=0.8128 params={'learning_rate': 0.000483907016003038, 'steps_per_switch': 7464, 'total_timesteps': 70587}
[2026-04-16 07:01:30] UCB=3.3445 mu=1.9911 σ=0.6767 params={'learning_rate': 0.00041395529913323, 'steps_per_switch': 7143, 'total_timesteps': 72719}
[2026-04-16 07:01:30] UCB=3.3380 mu=1.6221 σ=0.8579 params={'learning_rate': 0.0004908056727355321, 'steps_per_switch': 4967, 'total_timesteps': 69146}
[2026-04-16 07:01:30] [Wave4] Proposed params: {'learning_rate': 0.00033896311195395496, 'steps_per_switch': 5448, 'total_timesteps': 69359}
[2026-04-16 07:01:32] [Wave4] Launching trial 14: {'learning_rate': 0.00033896311195395496, 'steps_per_switch': 5448, 'total_timesteps': 69359}
[2026-04-16 07:01:32] [Wave4] Command: python3 /home/paulh/projects/donkeycar-rl-autoresearch/agent/multitrack_runner.py --total-timesteps 69359 --steps-per-switch 5448 --learning-rate 0.00033896311195395496 --eval-episodes 3 --save-dir /home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0014
[2026-04-16 07:54:33] [Wave4] Trial 14 finished in 3181.0s, rc=0
[2026-04-16 07:54:33] [Wave4] Parsed: combined=1573.0703 mini_monaco=1573.0703
[2026-04-16 07:54:33] [Wave4] ===== Trial 14 Summary =====
[2026-04-16 07:54:33] GP data points : 14
[2026-04-16 07:54:33] Wave4 Champion: trial=3 score=1943.10 params={'learning_rate': 0.0006852550685205609, 'steps_per_switch': 17499, 'total_timesteps': 157743}
[2026-04-16 07:54:33] Top 5:
[2026-04-16 07:54:33] score=1943.10 params={'learning_rate': 0.0006852550685205609, 'steps_per_switch': 17499, 'total_timesteps': 90000}
[2026-04-16 07:54:33] score=1573.07 params={'learning_rate': 0.00033896311195395496, 'steps_per_switch': 5448, 'total_timesteps': 69359}
[2026-04-16 07:54:33] score=1435.04 params={'learning_rate': 0.0007252855740444645, 'steps_per_switch': 6851, 'total_timesteps': 89893}
[2026-04-16 07:54:33] score=176.67 params={'learning_rate': 0.0016223486895735558, 'steps_per_switch': 5524, 'total_timesteps': 79532}
[2026-04-16 07:54:33] score=141.35 params={'learning_rate': 0.0010578566658808422, 'steps_per_switch': 4587, 'total_timesteps': 89501}
[2026-04-16 07:54:35]
[Wave4] ========== Trial 15/25 ==========
[2026-04-16 07:54:35] [Wave4] GP UCB top-5 proposals:
[2026-04-16 07:54:35] UCB=2.8282 mu=1.6797 σ=0.5742 params={'learning_rate': 0.0003994526493374, 'steps_per_switch': 7747, 'total_timesteps': 70559}
[2026-04-16 07:54:35] UCB=2.7924 mu=1.4316 σ=0.6804 params={'learning_rate': 0.0005427837302668084, 'steps_per_switch': 8198, 'total_timesteps': 71093}
[2026-04-16 07:54:35] UCB=2.7732 mu=1.5025 σ=0.6353 params={'learning_rate': 0.000186376586240458, 'steps_per_switch': 7967, 'total_timesteps': 71123}
[2026-04-16 07:54:35] UCB=2.7719 mu=1.2938 σ=0.7390 params={'learning_rate': 0.0002182534990992253, 'steps_per_switch': 8697, 'total_timesteps': 71541}
[2026-04-16 07:54:35] UCB=2.7193 mu=1.1369 σ=0.7912 params={'learning_rate': 0.00011645020756444703, 'steps_per_switch': 8822, 'total_timesteps': 69675}
[2026-04-16 07:54:35] [Wave4] Proposed params: {'learning_rate': 0.0003994526493374, 'steps_per_switch': 7747, 'total_timesteps': 70559}
[2026-04-16 07:54:37] [Wave4] Launching trial 15: {'learning_rate': 0.0003994526493374, 'steps_per_switch': 7747, 'total_timesteps': 70559}
[2026-04-16 07:54:37] [Wave4] Command: python3 /home/paulh/projects/donkeycar-rl-autoresearch/agent/multitrack_runner.py --total-timesteps 70559 --steps-per-switch 7747 --learning-rate 0.0003994526493374 --eval-episodes 3 --save-dir /home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0015
[2026-04-16 08:43:17] [Wave4] Trial 15 finished in 2920.0s, rc=0
[2026-04-16 08:43:17] [Wave4] Parsed: combined=111.2516 mini_monaco=111.2516
[2026-04-16 08:43:17] [Wave4] ===== Trial 15 Summary =====
[2026-04-16 08:43:17] GP data points : 15
[2026-04-16 08:43:17] Wave4 Champion: trial=3 score=1943.10 params={'learning_rate': 0.0006852550685205609, 'steps_per_switch': 17499, 'total_timesteps': 157743}
[2026-04-16 08:43:17] Top 5:
[2026-04-16 08:43:17] score=1943.10 params={'learning_rate': 0.0006852550685205609, 'steps_per_switch': 17499, 'total_timesteps': 90000}
[2026-04-16 08:43:17] score=1573.07 params={'learning_rate': 0.00033896311195395496, 'steps_per_switch': 5448, 'total_timesteps': 69359}
[2026-04-16 08:43:17] score=1435.04 params={'learning_rate': 0.0007252855740444645, 'steps_per_switch': 6851, 'total_timesteps': 89893}
[2026-04-16 08:43:17] score=176.67 params={'learning_rate': 0.0016223486895735558, 'steps_per_switch': 5524, 'total_timesteps': 79532}
[2026-04-16 08:43:17] score=141.35 params={'learning_rate': 0.0010578566658808422, 'steps_per_switch': 4587, 'total_timesteps': 89501}

5
agent/outerloop-results/autoresearch_results_phase4.jsonl
View File

@ -8,3 +8,8 @@
{"trial": 8, "timestamp": "2026-04-16T00:48:42.336768", "params": {"learning_rate": 0.0008478421211922181, "steps_per_switch": 14326, "total_timesteps": 89798}, "combined_test_score": 116.2588, "mini_monaco_reward": 116.2588, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0008/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 3762.1077337265015} {"trial": 8, "timestamp": "2026-04-16T00:48:42.336768", "params": {"learning_rate": 0.0008478421211922181, "steps_per_switch": 14326, "total_timesteps": 89798}, "combined_test_score": 116.2588, "mini_monaco_reward": 116.2588, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0008/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 3762.1077337265015}
{"trial": 9, "timestamp": "2026-04-16T02:11:13.466663", "params": {"learning_rate": 0.0007252855740444645, "steps_per_switch": 6851, "total_timesteps": 89893}, "combined_test_score": 1435.0401, "mini_monaco_reward": 1435.0401, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0009/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 4947.0997133255005} {"trial": 9, "timestamp": "2026-04-16T02:11:13.466663", "params": {"learning_rate": 0.0007252855740444645, "steps_per_switch": 6851, "total_timesteps": 89893}, "combined_test_score": 1435.0401, "mini_monaco_reward": 1435.0401, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0009/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 4947.0997133255005}
{"trial": 10, "timestamp": "2026-04-16T03:31:41.385487", "params": {"learning_rate": 0.0010578566658808422, "steps_per_switch": 4587, "total_timesteps": 89501}, "combined_test_score": 141.3547, "mini_monaco_reward": 141.3547, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0010/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 4823.869568347931} {"trial": 10, "timestamp": "2026-04-16T03:31:41.385487", "params": {"learning_rate": 0.0010578566658808422, "steps_per_switch": 4587, "total_timesteps": 89501}, "combined_test_score": 141.3547, "mini_monaco_reward": 141.3547, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0010/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 4823.869568347931}
{"trial": 11, "timestamp": "2026-04-16T04:47:26.275980", "params": {"learning_rate": 0.0004453206607211525, "steps_per_switch": 6345, "total_timesteps": 89491}, "combined_test_score": 85.169, "mini_monaco_reward": 85.169, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0011/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 4539.4217319488525}
{"trial": 12, "timestamp": "2026-04-16T06:01:18.667114", "params": {"learning_rate": 0.0008597472207162354, "steps_per_switch": 6936, "total_timesteps": 89806}, "combined_test_score": 131.9914, "mini_monaco_reward": 131.9914, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0012/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 4428.36554312706}
{"trial": 13, "timestamp": "2026-04-16T07:01:28.371417", "params": {"learning_rate": 0.0019124682919611855, "steps_per_switch": 3574, "total_timesteps": 73915}, "combined_test_score": 86.8189, "mini_monaco_reward": 86.8189, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0013/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 3605.6811096668243}
{"trial": 14, "timestamp": "2026-04-16T07:54:33.441210", "params": {"learning_rate": 0.00033896311195395496, "steps_per_switch": 5448, "total_timesteps": 69359}, "combined_test_score": 1573.0703, "mini_monaco_reward": 1573.0703, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0014/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 3181.0467441082}
{"trial": 15, "timestamp": "2026-04-16T08:43:17.443930", "params": {"learning_rate": 0.0003994526493374, "steps_per_switch": 7747, "total_timesteps": 70559}, "combined_test_score": 111.2516, "mini_monaco_reward": 111.2516, "model_path": "/home/paulh/projects/donkeycar-rl-autoresearch/agent/models/wave4-trial-0015/model.zip", "champion": false, "run_status": "ok", "elapsed_sec": 2919.980232000351}