From c45959245d44625687c3dd7f3c92fd388ca2e623 Mon Sep 17 00:00:00 2001
From: Haizhong <hzzheng@umich.edu>
Date: Mon, 25 May 2026 13:54:27 -0400
Subject: [PATCH 1/2] chore: add default CODEOWNERS

---
 .github/CODEOWNERS | 2 ++
 1 file changed, 2 insertions(+)
 create mode 100644 .github/CODEOWNERS

diff --git a/.github/CODEOWNERS b/.github/CODEOWNERS
new file mode 100644
index 0000000..0e93827
--- /dev/null
+++ b/.github/CODEOWNERS
@@ -0,0 +1,2 @@
+# Default owner for all paths.
+*   @haizhongzheng

From 5dad9137b79e5ec73c15dcc793d7ec4c27ee6a0b Mon Sep 17 00:00:00 2001
From: Haizhong <hzzheng@umich.edu>
Date: Mon, 25 May 2026 13:54:45 -0400
Subject: [PATCH 2/2] fix: forward extra CLI args to launch_trainer.py in 3
 recipe scripts

The two search recipes and the codegen-verifier multi-agent recipe were
missing the "$@" passthrough on their launch_trainer.py invocations, so
any overrides passed to the recipe script (e.g. total_train_steps=30,
stats_logger.wandb.name=...) were silently dropped before reaching the
trainer. The other 31 recipe scripts under examples/ already had this;
this brings the three holdouts in line.
---
 .../scripts/run_qwen3-8b-codegen-verifier-m2po-full-2node.sh    | 2 ++
 .../scripts/run_qwen2.5-7b-instruct-m2po-delta.sh               | 1 +
 .../scripts/run_qwen2.5-7b-instruct-m2po-full.sh                | 1 +
 3 files changed, 4 insertions(+)

diff --git a/examples/code-multi-agent/qwen3-8b-codegen-verifier-m2po-full-2node/scripts/run_qwen3-8b-codegen-verifier-m2po-full-2node.sh b/examples/code-multi-agent/qwen3-8b-codegen-verifier-m2po-full-2node/scripts/run_qwen3-8b-codegen-verifier-m2po-full-2node.sh
index 1224ba2..2db8220 100755
--- a/examples/code-multi-agent/qwen3-8b-codegen-verifier-m2po-full-2node/scripts/run_qwen3-8b-codegen-verifier-m2po-full-2node.sh
+++ b/examples/code-multi-agent/qwen3-8b-codegen-verifier-m2po-full-2node/scripts/run_qwen3-8b-codegen-verifier-m2po-full-2node.sh
@@ -146,6 +146,7 @@ WEIGHT_TRANSFER_HTTP_PORT="${WEIGHT_TRANSFER_HTTP_PORT_MODEL0}" \
     examples/launch_trainer.py \
     --config "${EXPERIMENT_CONFIG}" \
     --trainer trainer_model0 \
+    "$@" \
     2>&1 | tee "${LOG_DIR}/trainer_model0.log" &
 sleep 3
 
@@ -158,4 +159,5 @@ WEIGHT_TRANSFER_HTTP_PORT="${WEIGHT_TRANSFER_HTTP_PORT_MODEL1}" \
     examples/launch_trainer.py \
     --config "${EXPERIMENT_CONFIG}" \
     --trainer trainer_model1 \
+    "$@" \
     2>&1 | tee "${LOG_DIR}/trainer_model1.log"
diff --git a/examples/search/qwen2.5-7b-instruct-m2po-delta/scripts/run_qwen2.5-7b-instruct-m2po-delta.sh b/examples/search/qwen2.5-7b-instruct-m2po-delta/scripts/run_qwen2.5-7b-instruct-m2po-delta.sh
index 706009b..629b89f 100755
--- a/examples/search/qwen2.5-7b-instruct-m2po-delta/scripts/run_qwen2.5-7b-instruct-m2po-delta.sh
+++ b/examples/search/qwen2.5-7b-instruct-m2po-delta/scripts/run_qwen2.5-7b-instruct-m2po-delta.sh
@@ -134,4 +134,5 @@ WEIGHT_TRANSFER_HTTP_PORT="${WEIGHT_TRANSFER_HTTP_PORT_MODEL0}" \
     examples/launch_trainer.py \
     --config "${EXPERIMENT_CONFIG}" \
     --trainer trainer_model0 \
+    "$@" \
     2>&1 | tee "${LOG_DIR}/trainer_model0.log"
diff --git a/examples/search/qwen2.5-7b-instruct-m2po-full/scripts/run_qwen2.5-7b-instruct-m2po-full.sh b/examples/search/qwen2.5-7b-instruct-m2po-full/scripts/run_qwen2.5-7b-instruct-m2po-full.sh
index 318bffb..66fdc06 100755
--- a/examples/search/qwen2.5-7b-instruct-m2po-full/scripts/run_qwen2.5-7b-instruct-m2po-full.sh
+++ b/examples/search/qwen2.5-7b-instruct-m2po-full/scripts/run_qwen2.5-7b-instruct-m2po-full.sh
@@ -134,4 +134,5 @@ WEIGHT_TRANSFER_HTTP_PORT="${WEIGHT_TRANSFER_HTTP_PORT_MODEL0}" \
     examples/launch_trainer.py \
     --config "${EXPERIMENT_CONFIG}" \
     --trainer trainer_model0 \
+    "$@" \
     2>&1 | tee "${LOG_DIR}/trainer_model0.log"