From 290201f62ec938b9b6c66ea7c81482a42a7e8631 Mon Sep 17 00:00:00 2001
From: Roland Tannous <rolandtannous@gravityq.ai>
Date: Thu, 28 May 2026 13:39:13 +0400
Subject: [PATCH] Studio: trim captioner logs to invoked+complete, render
 subprocess logs as JSON

Two changes to the RAG captioning log output:

  - Drop the noisy per-image and path-selection info lines
    (using-chat-VLM, loading-helper, per-image done). Only the
    'caption_images: invoked' and 'caption_images: complete' lines
    remain; warnings for genuine failures (helper load, per-image
    request, helper unload) are kept.
  - Configure structlog at the top of the ingestion subprocess worker
    with the same env the parent uses. The worker runs in a spawned
    process where structlog was never set up, so its logs fell back to
    structlog's dev ConsoleRenderer ([info] ...) instead of the JSON
    renderer the rest of the app uses. Now captioner/parser logs from
    the subprocess match the parent's JSON format.
---
 studio/backend/core/rag/captioner.py | 16 +---------------
 studio/backend/core/rag/ingestion.py | 14 ++++++++++++++
 2 files changed, 15 insertions(+), 15 deletions(-)

diff --git a/studio/backend/core/rag/captioner.py b/studio/backend/core/rag/captioner.py
index 857d8e54c6..b488bd7dcd 100644
--- a/studio/backend/core/rag/captioner.py
+++ b/studio/backend/core/rag/captioner.py
@@ -170,13 +170,7 @@ def caption_images(
         if vlm_url and vlm_model:
             endpoint = f"{vlm_url.rstrip('/')}/v1/chat/completions"
             model_name = vlm_model
-            logger.info(
-                "caption_images: using loaded chat VLM",
-                endpoint = endpoint,
-                model = model_name,
-            )
         else:
-            logger.info("caption_images: chat VLM unavailable, loading helper")
             loaded = _load_helper_vlm()
             if loaded is None:
                 logger.warning(
@@ -186,20 +180,12 @@ def caption_images(
             helper_backend, helper_base_url, helper_model_name = loaded
             endpoint = f"{helper_base_url.rstrip('/')}/v1/chat/completions"
             model_name = helper_model_name
-            logger.info("caption_images: using helper VLM", endpoint = endpoint)
 
         out: list[str] = []
         with httpx.Client(timeout = _REQUEST_TIMEOUT_SECONDS) as client:
             for idx, blob in enumerate(image_bytes_list):
                 try:
-                    caption = _post_one(client, endpoint, model_name, blob)
-                    out.append(caption)
-                    logger.info(
-                        "caption_images: per-image done",
-                        idx = idx,
-                        caption_len = len(caption),
-                        preview = caption[:80],
-                    )
+                    out.append(_post_one(client, endpoint, model_name, blob))
                 except Exception as exc:  # noqa: BLE001
                     logger.warning(
                         "caption_images: per-image request failed",
diff --git a/studio/backend/core/rag/ingestion.py b/studio/backend/core/rag/ingestion.py
index f8c286d165..32614a8e5f 100644
--- a/studio/backend/core/rag/ingestion.py
+++ b/studio/backend/core/rag/ingestion.py
@@ -64,6 +64,20 @@ def _subprocess_worker(
     vlm_url: str | None = None,
     vlm_model: str | None = None,
 ) -> None:
+    # Spawned subprocess: structlog isn't configured here (the parent's
+    # setup runs in the FastAPI process only), so configure it the same
+    # way so captioner / parser logs render as JSON like the rest, not
+    # structlog's default dev ConsoleRenderer.
+    try:
+        import os as _os
+
+        from loggers.config import LogConfig
+
+        LogConfig.setup_logging(
+            env = _os.getenv("ENVIRONMENT_TYPE", "production"),
+        )
+    except Exception:  # noqa: BLE001
+        pass
     try:
         from core.rag.captioner import caption_images
         from core.rag.chunking import chunk_pages