getting timeout error when using gpt5
# helpm
manish_iitg
09/02/2025, 3:56 PMhi, i am using gpt5 with medium reasoning.. this produces a very large output.. and i am getting this error BamlClientHttpError(client_name=Customgpt5Client, message=LLM call failed: LLMErrorResponse { client: "Customgpt5Client", model: None, prompt: Chat([RenderedChatMessage { role: "system", allow_duplicate_role: false, parts: [Text("You will be provided a list of \"inferred\" OKR JSON objects and employee narrative objects. ..... [very long input in between]RenderedChatMessage { role: "user", allow_duplicate_role: false, parts: [Text("CONFIRMED. Generate the complete, revised org_wide_okr_model following the schema provided.")] }]), request_options: {"model": String("gpt-5"), "reasoning_effort": String("minimal")}, start_time: SystemTime { tv_sec: 1756514867, tv_nsec: 905846629 }, latency: 900.199564937s, message: "Failed to parse JSON: expected value at line 1 column 1", code: Other(504) }, status_code=504)
manish_iitg
09/02/2025, 3:56 PMis there some place to increase time where is 900sec timeout configured
manish_iitg
09/02/2025, 3:59 PM@Vaibhav (BAML 🐑) hi, if you can help out
manish_iitg
09/02/2025, 3:59 PMi am working with halovision
v
Vaibhav (BAML 🐑)
09/02/2025, 4:28 PMhi @manish_iitg thanks for the context! I wonder if @baml-sam or @aaronv (BAML) can help out here. I'm out on vacation this week so delayed respones on my end! 🙂
b
baml-sam
09/02/2025, 9:35 PMHmm - we don't have any default timeout configured
baml-sam
09/02/2025, 9:37 PMSo I'm not sure where that 900s timeout is coming from
baml-sam
09/02/2025, 9:37 PMAlso, have you tried debugging with the raw curl view? We have a tab in the playground that allows you see the curl request that we send to the model on your behalf
m
manish_iitg
09/03/2025, 3:38 AMneed to try that . will test in playground
j
JJJasper
09/03/2025, 3:37 PMlooks like the model is set to None - how is your "Customgpt5Client" client set up?