From 1e2238547962ac9cb428232c417b39c114b605fb Mon Sep 17 00:00:00 2001 From: Cursor Agent Date: Thu, 19 Mar 2026 02:43:06 +0000 Subject: [PATCH] fix(streaming): handle RuntimeError during model_copy in streaming handler The race condition occurs when model_copy(deep=True) tries to deepcopy _hidden_params dict while it's being concurrently modified by logging callbacks. Fall back to shallow copy if the deep copy fails. Co-authored-by: Ishaan Jaff --- .../litellm_core_utils/streaming_handler.py | 26 ++++++++++++++----- 1 file changed, 19 insertions(+), 7 deletions(-) diff --git a/litellm/litellm_core_utils/streaming_handler.py b/litellm/litellm_core_utils/streaming_handler.py index 6e991e6911b..ca78e72c693 100644 --- a/litellm/litellm_core_utils/streaming_handler.py +++ b/litellm/litellm_core_utils/streaming_handler.py @@ -1893,15 +1893,23 @@ class CustomStreamWrapper: "usage", getattr(complete_streaming_response, "usage"), ) - self.cache_streaming_response( - processed_chunk=complete_streaming_response.model_copy( + try: + _cache_copy = complete_streaming_response.model_copy( deep=True - ), + ) + _log_copy = complete_streaming_response.model_copy( + deep=True + ) + except RuntimeError: + _cache_copy = complete_streaming_response.model_copy() + _log_copy = complete_streaming_response.model_copy() + self.cache_streaming_response( + processed_chunk=_cache_copy, cache_hit=cache_hit, ) executor.submit( self.logging_obj.success_handler, - complete_streaming_response.model_copy(deep=True), + _log_copy, None, None, cache_hit, @@ -2113,11 +2121,15 @@ class CustomStreamWrapper: "usage", getattr(complete_streaming_response, "usage"), ) + try: + _copy = complete_streaming_response.model_copy( + deep=True + ) + except RuntimeError: + _copy = complete_streaming_response.model_copy() asyncio.create_task( self.async_cache_streaming_response( - processed_chunk=complete_streaming_response.model_copy( - deep=True - ), + processed_chunk=_copy, cache_hit=cache_hit, ) )