langfuse generation spans with token usage; switch llm to vllm (qwen2.5-coder-7b on mcrndeb)

api/analyses/compare_periods.py

@@ -83,7 +83,10 @@ class ComparePeriods(Analysis):
                     system_len=len(interpret_system),
                     user_len=len(interpret_user),
                 ))
-                summary = chat(system=interpret_system, user=interpret_user, max_tokens=512)
+                summary = chat(
+                    system=interpret_system, user=interpret_user,
+                    max_tokens=512, span_name="compare_periods.interpret",
+                )
             except Exception as e:
                 logger.exception("compare_periods failed")
                 await events.publish_current(events.tool_call_end("compare_periods", error=str(e)))
@@ -122,6 +125,7 @@ def _generate_pair(question: str, period_a: str, period_b: str) -> dict[str, dic
             metrics_block=schema.render_metrics(),
         ),
         max_tokens=1024,
+        span_name="compare_periods.pair",
     )
     obj = json.loads(_extract_json(text))
     for k in ("a", "b"):