Langfuse

Langfuse is the open source LLM Engineering Platform. This is our Community Discord.

Join

Langfuse

Langfuse is the open source LLM Engineering Platform. This is our Community Discord.

Join

support

self-host-support

get-support

feedback

feature-suggestion

self-host-discussion

announcements-releases

Bhasker

4/18/2024

which version are you using? run_id not

which version are you using? run_id not found is fixed last week

byron mackay

4/15/2024

My local instance of langfuse is showing

My local instance of langfuse is showing null for input, output, and metadata on the top trace. If I dig down, I see the outputs showing. I'm using a simple LangChain implementation to run an LLM (code in thread). Am I missing something?

byron mackay

4/10/2024

I'm having issues with the `api/public/

UPDATE: Issue resolved. Didn't run the DB migrations. I'm having issues with the api/public/dataset-run-items endpoint not returning trace_ids. DataRunItem requires a trace_id, but they don't appear in the response. Stacktrace in thread. Any ideas on how to fix this?...

Nikolass

4/9/2024

n8n

Hello, I'm discovering Langfuse, and I'd like to know if I can use it with n8n (as an http node). n8n will then handle the calls to the models , then responses would be redirected to langfuse for evaluating ?...

AdrianC

3/22/2024

Thread

mkern

3/20/2024

Seeing a lot of httpx timeouts when

Seeing a lot of httpx timeouts when calling langfuse_callback_handler.flush() with llama index

message.txt

Bill

3/13/2024

View traces

Is there a way to view all traces in a project? It appears that you have to select specific users to actually see traces.

denis56

3/12/2024

Dev/staging/prod

Hello team! I just wanted to double check about best practice for separating local/dev/prod environments in Langfuse. As far as understand, the recommended approach is to use tags for that. Would creating a separate project make sense too, would like to understand some pros and cons regarding this approach

sumanth balaji

3/5/2024

DB connections aren’t reused when self hosting Langfuse

on further inspection, this sharp increase of db connections from langfuse instance has occured in the past too. Only now have we realised the issue due to error logs

Ryan Ribeiro

2/26/2024

hello, I have an application in which I

hello, I have an application in which I am building a base class for a task, this application will relate to an llm and can use both langchain and llamaindex, and I would like to have a lot of control over the langfuse traces so I need to go down the level. For example, a problem would be the langchain chains in which they concentrate a runnablesequence and the invoke triggers this sequence, but how would I treat the spans in this case since the processes occur hidden? Is the best way to not use...

aazizi

2/26/2024

Python sdk issue

Hello, I require assistance folk, I encountered an error this morning.: <ipython-input-5-c6633758d278> in get_traces(name, limit, user_id, order_by) 4 5 while True: ----> 6 response = langfuse.client.trace.list( 7 name=name, page=page, user_id=user_id, order_by=order_by...

sumanth balaji

2/12/2024

the following is the logs when debug=

the following is the logs when debug=True is added during the langfuse object instantiation in the code snippet Scores still are not received by the langfuse server. `Evaluating: 100%|██████████| 2/2 [00:09<00:00, 4.69s/it]240a5f73e468d5407fd) [2024-02-12T15:10:05+05:30] (ecs/nlq-worker/15e524848f864240a5f73e468d5407fd) DEBUG:langfuse:Creating score {'id': '5c8c6162-4e07-4b9a-b12c-0f632d93fc1f', 'trace_id': '8885ab98-9d0e-446e-9382-7e7ee23c697d', 'observation_id': None, 'name': 'faithfulness', 'value': 1.0, 'comment': None}......

Marc

2/6/2024

Hi Denis, currently redaction does not

Hi Denis, currently redaction does not work well if you aim for the automated usage/cost analysis in Langfuse. Do you have access to the usage unit (eg tokens)? Then it'd be easy to still retain price calculations in Langfuse while redacting some of the input/output. Especially enteprise teams do this frequently for production environments

Aamir Rasheed

2/5/2024

Hey all! I'm using Langfuse with LiteLLM

Hey all! I'm using Langfuse with LiteLLM (using success_callback set to langfuse) and I am getting the following error: langfuse.request.APIError: {'message': 'Invalid public key'} (401): None Does anyone know why this is happening?...

kanundrum

1/26/2024

Any pointers as to what could cause this

Any pointers as to what could cause this odd behaviour? (it's happening when I point to the hosted solution or self hosted)

Marc

1/24/2024

Good question, haven't used the flowise

Good question, haven't used the flowise integration myself in prod though. What you could do is to fetch the session via the GET API which returns the traceids that belong to the session

lephant_

1/24/2024

LCEL streaming

Hi all, I'm using langchain LCEL + stream, so chain.stream... But now the inputs at on_chain_start are initially empty {inputs:''} ```FIRST: -------> {'input': ''} FIRST: -------> {'input': ''} FIRST: -------> {'input': ''}...

Harpinder

1/24/2024

Hey folks, do we have some restrictions

Hey folks, do we have some restrictions on what kind of input/outputs a "chain" can have in LangChain while integrating with Langfuse? I was trying Langfuse, but got an error

Exception in adding task Object of type AnswerChainInput is not JSON serializable

Exception in adding task Object of type AnswerChainInput is not JSON serializable

where, ```...

Raji

1/23/2024

Scores & Evaluation - Langfuse

Hi, I have a query regarding scores related to "similarity to prompt injections" and "refusals" as mentioned in https://langfuse.com/docs/scores under "Kinds of scores". However none of the attributes from langchain eval or ragas attributes address these two. Have any of you implemented scores to address these aspects?

pavelf

1/22/2024

Different traces

HI! I am trying to use feedback in the same trace that langchain callback handler creates. I followed https://langfuse.com/docs/langchain/python#adding-scores and ran into a problem: It generates two empty extra traces as well as it puts the feedback into another, fourth one. I will also provide all relevant code in the thread

Previous Next