langchain-ai/langserve
1
0
Fork 0
mirror of https://github.com/langchain-ai/langserve.git synced 2026-07-16 09:34:30 -04:00
Code Issues 136 Packages Projects Releases 10 Wiki Activity

Clientside passed Callbackhandler not triggered when streaming on LangserveRunnable #191

New Issue
Open
opened 2026-02-16 00:19:17 -05:00 by yindo · 0 comments
yindo commented 2026-02-16 00:19:17 -05:00
Owner
Copy Link
Copy Source

Originally created by @DJ2695 on GitHub (Jun 10, 2024).

Me and some colleagues observed an issue, where Langchain Callback Handlers get not triggered if passed as callbacks on Clientside within the Streaming function. The Runnable contains RAG Components like a Retriever. The chain/runnable was tested without Langserve and the callbacks were triggered (on_retriever_end, on_new_llm_token, on_chat_model_start) just as expected. Using Langserve's RemoteRunnable, this behaviour changed.

Are we missing out on some important information regarding clientside streaming on a RemoteRunnable or could this be an implementation Issue?

The following code shows, how it looks like


def azure_ai_chat():
    return RemoteRunnable("http://localhost:8001/azure-ai-search/")

# Langserve Runnable: Testes and Works as endpoint as well as without Langserve
runnable = azure_ai_chat()  # type: Runnable

ai_response = cl.Message(content="", elements=[], actions=[])
token_usage = TokenUsage()
async for chunk in runnable.astream(
    {"question": message.content},
    config=RunnableConfig(callbacks=[cl.LangchainCallbackHandler(),
                                     TokenUsageCallbackHandler(
                                         "gpt-4", token_usage),
                                     DocsMessageHandler(ai_response)
                                     ]),

):
    # Stream specific code
    if "answer" in chunk.keys():
                await response_msg.stream_token(chunk["answer"].content)
                
    if chunk and "context" in chunk.keys():
        relevant_docs = chunk["context"]
Originally created by @DJ2695 on GitHub (Jun 10, 2024). Me and some colleagues observed an issue, where Langchain Callback Handlers get not triggered if passed as callbacks on Clientside within the Streaming function. The Runnable contains RAG Components like a Retriever. The chain/runnable was tested without Langserve and the callbacks were triggered (on_retriever_end, on_new_llm_token, on_chat_model_start) just as expected. Using Langserve's RemoteRunnable, this behaviour changed. Are we missing out on some important information regarding clientside streaming on a RemoteRunnable or could this be an implementation Issue? The following code shows, how it looks like ``` python def azure_ai_chat(): return RemoteRunnable("http://localhost:8001/azure-ai-search/") # Langserve Runnable: Testes and Works as endpoint as well as without Langserve runnable = azure_ai_chat() # type: Runnable ai_response = cl.Message(content="", elements=[], actions=[]) token_usage = TokenUsage() async for chunk in runnable.astream( {"question": message.content}, config=RunnableConfig(callbacks=[cl.LangchainCallbackHandler(), TokenUsageCallbackHandler( "gpt-4", token_usage), DocsMessageHandler(ai_response) ]), ): # Stream specific code if "answer" in chunk.keys(): await response_msg.stream_token(chunk["answer"].content) if chunk and "context" in chunk.keys(): relevant_docs = chunk["context"] ```
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
bug
confirmed
duplicate
enhancement
help wanted
in-progress
investigate
not langserve
playground
pull-request
Mirrored from GitHub Pull Request
 question
stale
streaming
unable-to-reproduce
wontfix
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langchain-ai/langserve#191
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?