Find answers from the community

Updated 2 months ago

Hi good afternoon,

Hi good afternoon,
My flask api kicks tasks to a redis queue. Workers grab the tasks and run the jobs. This works great, but for the multistep query engine queries when deployed they time out usually and takes much longer (9mins +). Locally / docker desktop these queries complete much faster (under 2 minutes), and so yeah only when deployed getting hung up. I kicked up the specs for the deployment to 8gb of memory and 4 cpu and still same issue. Before i start changing this code (not originally my code) wondering if any other ideas or best to just keep upping the specs?
n
1 comment
https://cloud.google.com/run/docs/configuring/services/memory-limits

query_engine = MultiStepQueryEngine(
query_engine=single_step_query_engine,
query_transform=step_decompose_transform,
index_summary="Ask me anything.",
num_steps=6,
)

Looks like deployment is "todo" https://docs.llamaindex.ai/en/stable/understanding/deployment/deployment/
Add a reply
Sign up and join the conversation on Discord