Batching Fix (#474)

python/langsmith/client.py

@@ -1134,6 +1134,7 @@ def batch_ingest_runs(
                 if body_size > 0 and body_size + len(body[0]) > size_limit_bytes:
                     self._post_batch_ingest_runs(orjson.dumps(body_chunks))
                     body_size = 0
+                    body_chunks.clear()
                 body_size += len(body[0])
                 body_chunks[key].append(orjson.Fragment(body.popleft()))
         if body_size:

python/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "langsmith"
-version = "0.1.7"
+version = "0.1.8"
 description = "Client library to connect to the LangSmith LLM Tracing and Evaluation Platform."
 authors = ["LangChain <[email protected]>"]
 license = "MIT"

python/tests/unit_tests/test_client.py

@@ -803,11 +803,6 @@ def test_batch_ingest_run_splits_large_batches(payload_size: int):
         sum([1 for call in mock_session.request.call_args_list if call[0][0] == "post"])
         == expected_num_requests
     )
-
-    # Use orjson to reload all the post request bodies and check that all the
-    # post/patch requests are present in the request bodies
-    import orjson
-
     request_bodies = [
         op
         for call in mock_session.request.call_args_list
@@ -819,3 +814,6 @@ def test_batch_ingest_run_splits_large_batches(payload_size: int):
     # Check that all the run_ids are present in the request bodies
     for run_id in all_run_ids:
         assert any([body["id"] == str(run_id) for body in request_bodies])
+
+    # Check that no duplicate run_ids are present in the request bodies
+    assert len(request_bodies) == len(set([body["id"] for body in request_bodies]))