How to increase waiting time?

At a glance

The community member is using a local LLM model from ollama + qdrant and is experiencing timeout errors when querying the model. They have confirmed that the ollama server is running and they can access it, and they can also chat with the model using the terminal. However, they are still encountering the same timeout errors. Another community member suggests setting the request_timeout parameter to 30 seconds, which is the default value.

ppikachu8887867

How to increase waiting time?

I'm using a local LLM model from ollama + qdrant:

Plain Text

llm = Ollama(model="model")
...
query_engine = index.as_query_engine()
response = query_engine.query("query")

but keep getting:

Plain Text

TimeoutError: timed out
...
httpcore.ReadTimeout: timed out
...
httpx.ReadTimeout: timed out

I confirm that ollama server is running and I can access: http://localhost:11434

I also confirm that I can chat with the model using terminal with ollama run command.

I played around with the raw llm (without qdrant) and it worked, even though sometimes it used to throw the same error.

3 comments

LLogan M

llm = Ollama(model="model", request_timeout=30)

LLogan M

30s is the default

ppikachu8887867

@Logan M Ah shoot! Thank you!

Add a reply

Find answers from the community

How to increase waiting time?