From 2c9002785ef04dbf72315028761e9c8878382f36 Mon Sep 17 00:00:00 2001
From: Codex <codex@geointel.local>
Date: Tue, 7 Jul 2026 05:21:54 +0200
Subject: [PATCH] Harden operator YOLO training runtime

---
 backend/README.md                       | 3 ++-
 docs/AI_PIPELINES.md                    | 3 ++-
 docs/CODEX_EXECUTION_LOG.md             | 1 +
 scripts/README.md                       | 3 ++-
 scripts/train_operator_yolo_detector.sh | 5 ++++-
 5 files changed, 11 insertions(+), 4 deletions(-)

diff --git a/backend/README.md b/backend/README.md
index c4ad8f9a..072dca3d 100644
--- a/backend/README.md
+++ b/backend/README.md
@@ -283,7 +283,7 @@ candidates are too weak for the target imagery. It is not a browser feature and
 does not change API contracts:
 
 ```bash
-docker exec -it geointel python /app/scripts/export_operator_yolo_dataset.py \
+docker exec -it geointel python3 /app/scripts/export_operator_yolo_dataset.py \
   --manifest-path /app/storage/operator-data/operator_samples_manifest.json \
   --output-dir /app/storage/operator-data/yolo-building-dataset \
   --val-samples turnhout \
@@ -302,6 +302,7 @@ docker exec \
   -e TRAIN_BATCH=2 \
   -e TRAIN_WORKERS=0 \
   -e TRAIN_DEVICE=cpu \
+  -e PYTHON_BIN=python3 \
   geointel bash /app/scripts/train_operator_yolo_detector.sh
 ```
 
diff --git a/docs/AI_PIPELINES.md b/docs/AI_PIPELINES.md
index 32a51a88..212ba7aa 100644
--- a/docs/AI_PIPELINES.md
+++ b/docs/AI_PIPELINES.md
@@ -233,7 +233,7 @@ When repeated public model benchmarks remain too weak, the operator can convert
 the prepared real-data samples into a local YOLO training dataset:
 
 ```bash
-docker exec -it geointel python /app/scripts/export_operator_yolo_dataset.py \
+docker exec -it geointel python3 /app/scripts/export_operator_yolo_dataset.py \
   --manifest-path /app/storage/operator-data/operator_samples_manifest.json \
   --output-dir /app/storage/operator-data/yolo-building-dataset \
   --val-samples turnhout \
@@ -258,6 +258,7 @@ docker exec \
   -e TRAIN_BATCH=2 \
   -e TRAIN_WORKERS=0 \
   -e TRAIN_DEVICE=cpu \
+  -e PYTHON_BIN=python3 \
   geointel bash /app/scripts/train_operator_yolo_detector.sh
 ```
 
diff --git a/docs/CODEX_EXECUTION_LOG.md b/docs/CODEX_EXECUTION_LOG.md
index 89ededce..4294342a 100644
--- a/docs/CODEX_EXECUTION_LOG.md
+++ b/docs/CODEX_EXECUTION_LOG.md
@@ -9,6 +9,7 @@ Changed:
   - requires `OPERATOR_YOLO_DATASET_DIR`
   - requires an existing `YOLO_BASE_MODEL_PATH`
   - writes a local `TRAIN_MODEL_OUTPUT_PATH`
+  - uses `PYTHON_BIN=python3` by default for the all-in-one container
   - writes `training_summary.json`.
 - Added readiness coverage for exporter compile and train-wrapper shell syntax.
 - Added regression coverage in `backend/tests/test_sprint129_operator_yolo_training_dataset.py`.
diff --git a/scripts/README.md b/scripts/README.md
index fbda2b12..62c15fab 100644
--- a/scripts/README.md
+++ b/scripts/README.md
@@ -265,7 +265,7 @@ Export the same operator samples to a local YOLO detection dataset when the
 public model candidates are not strong enough for the target imagery:
 
 ```bash
-docker exec -it geointel python /app/scripts/export_operator_yolo_dataset.py \
+docker exec -it geointel python3 /app/scripts/export_operator_yolo_dataset.py \
   --manifest-path /app/storage/operator-data/operator_samples_manifest.json \
   --output-dir /app/storage/operator-data/yolo-building-dataset \
   --val-samples turnhout \
@@ -291,6 +291,7 @@ docker exec \
   -e TRAIN_BATCH=2 \
   -e TRAIN_WORKERS=0 \
   -e TRAIN_DEVICE=cpu \
+  -e PYTHON_BIN=python3 \
   geointel bash /app/scripts/train_operator_yolo_detector.sh
 ```
 
diff --git a/scripts/train_operator_yolo_detector.sh b/scripts/train_operator_yolo_detector.sh
index 409fe035..068c7a5c 100644
--- a/scripts/train_operator_yolo_detector.sh
+++ b/scripts/train_operator_yolo_detector.sh
@@ -21,6 +21,7 @@ Environment variables:
   TRAIN_BATCH                  Batch size. Default: 2
   TRAIN_WORKERS                Data-loader workers. Default: 0
   TRAIN_DEVICE                 Device passed to Ultralytics. Default: cpu
+  PYTHON_BIN                   Python executable. Default: python3
 
 This helper is an operator/runtime smoke wrapper. It requires an existing local
 base model and an existing local dataset.yaml. It does not create app features.
@@ -42,6 +43,7 @@ TRAIN_IMGSZ="${TRAIN_IMGSZ:-512}"
 TRAIN_BATCH="${TRAIN_BATCH:-2}"
 TRAIN_WORKERS="${TRAIN_WORKERS:-0}"
 TRAIN_DEVICE="${TRAIN_DEVICE:-cpu}"
+PYTHON_BIN="${PYTHON_BIN:-python3}"
 
 DATASET_YAML="${OPERATOR_YOLO_DATASET_DIR%/}/dataset.yaml"
 SUMMARY_PATH="${TRAIN_OUTPUT_DIR%/}/${TRAIN_RUN_NAME}/training_summary.json"
@@ -55,6 +57,7 @@ export TRAIN_IMGSZ
 export TRAIN_BATCH
 export TRAIN_WORKERS
 export TRAIN_DEVICE
+export PYTHON_BIN
 export SUMMARY_PATH
 
 if [[ ! -f "${DATASET_YAML}" ]]; then
@@ -69,7 +72,7 @@ fi
 
 mkdir -p "${TRAIN_OUTPUT_DIR}" "$(dirname "${TRAIN_MODEL_OUTPUT_PATH}")"
 
-python - <<'PY'
+"${PYTHON_BIN}" - <<'PY'
 from __future__ import annotations
 
 import json