lyra/summary.py

@@ -129,16 +129,20 @@ def maybe_summarize_async(session_id: str, backend: Backend | None = None) -> No
 
 
 def summarize_all(
-    backend: Backend | None = None, limit: int | None = None, workers: int = 8
+    backend: Backend | None = None, limit: int | None = None, workers: int | None = None
 ) -> dict:
     """Summarize every session that needs it. Idempotent and resumable.
 
-    LLM summarization runs concurrently across `workers` threads (great for a
+    Concurrency is backend-aware: the cloud API parallelizes happily, but the
-    cloud backend). DB reads (loading transcripts) and writes (store_summary,
+    local/MI50 GPU servers run a single slot (llama.cpp --parallel 1) — firing N
-    which also embeds) happen on the main thread, so the single SQLite
+    requests at them just queues, blows the client timeout, and thrashes the KV
-    connection is never touched from multiple threads.
+    cache (wasted compute + heat). So GPU backends run serially unless overridden.
+    DB reads/writes (store_summary embeds) stay on the main thread, so the single
+    SQLite connection is never touched from multiple threads.
     """
     backend = backend or config.load().summary_backend
+    if workers is None:
+        workers = 8 if backend == "cloud" else 1
 
     # Main thread: collect the work (transcripts) for sessions needing a summary.
     todo: list[tuple[str, str, int]] = []