pyr0ball e2658f743f feat(scheduler): OrchestratedScheduler for cloud/multi-GPU, configurable via env ... Switches to OrchestratedScheduler in cloud mode so concurrent recipe_llm jobs fan out across all registered cf-orch GPU nodes instead of serializing on one. Under load this eliminates poll timeouts from queue backup. USE_ORCH_SCHEDULER env var gives explicit control independent of CLOUD_MODE: unset follow CLOUD_MODE (cloud=orch, local=local) true OrchestratedScheduler always (e.g. multi-GPU local rig) false LocalScheduler always (e.g. cloud single-GPU dev instance) ImportError fallback: if circuitforge_orch is not installed and orch is requested, logs a warning and falls back to LocalScheduler gracefully.		2026-04-19 22:11:34 -07:00
..
__init__.py	feat(tasks): add background task scheduler for LLM expiry fallback	2026-03-31 09:25:48 -07:00
runner.py	feat(recipes): async L3/L4 recipe job queue with poll endpoint	2026-04-19 21:44:27 -07:00
scheduler.py	feat(scheduler): OrchestratedScheduler for cloud/multi-GPU, configurable via env	2026-04-19 22:11:34 -07:00