SirStone fedab54bc0 feat(PPO_Bot): multi-round transition accumulation (UPDATE_INTERVAL=10) ...

- Accumulate transitions across 10 rounds (~3000) before PPO update
  (was per-round ~300 — gradient estimates were far too noisy)
- training.nim: MAX_TRANSITIONS 4096→8192, done flag on transitions,
  GAE handles episode boundaries correctly
- PPO_Bot.nim: buffer persists across rounds, update every N rounds
- training.env: lr 5e-5→1e-4, entropy 0.001, UPDATE_INTERVAL=10

2026-08-20 15:06:00 +02:00

..

battle_runner

feat(PPO_Bot): command abstraction layer — goto/aimTo controllers (#24)

2026-08-17 19:20:40 +02:00

training_runner

feat(PPO_Bot): multi-round transition accumulation (UPDATE_INTERVAL=10)

2026-08-20 15:06:00 +02:00

warm_start.py

feat(PPO_Bot): bot-relative bullets + scan staleness (STATE_DIM=57)

2026-08-20 14:27:28 +02:00