From 91869c531c8f1e0e28090d69ad5e50a74c76b641 Mon Sep 17 00:00:00 2001 From: Gahow Wang Date: Wed, 15 Jul 2026 23:17:43 +0800 Subject: [PATCH] Add blind Qwen235B fleet job --- .../best_effort/fleet_fixed_cohort.toml | 25 +++++++++++++++++++ .../best_effort/jobs_fixed_cohort.toml | 10 ++++++++ 2 files changed, 35 insertions(+) create mode 100644 runs/frontier-multicase-sufficiency-v0/best_effort/fleet_fixed_cohort.toml create mode 100644 runs/frontier-multicase-sufficiency-v0/best_effort/jobs_fixed_cohort.toml diff --git a/runs/frontier-multicase-sufficiency-v0/best_effort/fleet_fixed_cohort.toml b/runs/frontier-multicase-sufficiency-v0/best_effort/fleet_fixed_cohort.toml new file mode 100644 index 0000000..16cfc56 --- /dev/null +++ b/runs/frontier-multicase-sufficiency-v0/best_effort/fleet_fixed_cohort.toml @@ -0,0 +1,25 @@ +version = 1 + +[paths] +state_dir = ".aituner/gpu_fleet/fixed_cohort_state" +artifacts_dir = ".aituner/gpu_fleet/fixed_cohort_artifacts" + +[ssh] +connect_timeout_sec = 10 + +[scheduler] +gpu_free_memory_mb = 1024 +gpu_free_utilization_pct = 10 +prefer_pack = true + +[sync] +mode = "rsync" +local_path = "runs/frontier-multicase-sufficiency-v0/best_effort" +exclude = ["__pycache__/", "*.pyc"] + +[[hosts]] +name = "dash0" +ssh_alias = "dash0" +enabled = true +sync_remote_path = "/home/admin/cpfs/wjh/aituner/aituner-frontier-best-effort-20260715/runs/frontier-multicase-sufficiency-v0/best_effort" +fleet_root = "/home/admin/.aituner_gpu_fleet_fixed_cohort" diff --git a/runs/frontier-multicase-sufficiency-v0/best_effort/jobs_fixed_cohort.toml b/runs/frontier-multicase-sufficiency-v0/best_effort/jobs_fixed_cohort.toml new file mode 100644 index 0000000..afd140d --- /dev/null +++ b/runs/frontier-multicase-sufficiency-v0/best_effort/jobs_fixed_cohort.toml @@ -0,0 +1,10 @@ +# Immutable one-job queue for the blind Qwen235B fixed-cohort rank experiment. +version = 1 + +[[jobs]] +name = "qwen235b-fixed-cohort-community-v1-blind-20260715" +gpus = 8 +gpu_model = "H20" +hosts = ["dash0"] +command = "cd ../../.. && /tmp/wjh-frontier-vllm0102-smoke/.venv/bin/python runs/frontier-multicase-sufficiency-v0/best_effort/fixed_cohort_rank.py run-community --manifest /home/admin/cpfs/wjh/frontier-community-qwen235-smoke-20260715/rankable-fixed-cohort-v1/community-v1-blind/run_manifest.json" +env = { PYTHONPATH = "/home/admin/cpfs/wjh/aituner/aituner-frontier-best-effort-20260715/src" }