Closed babushkai closed 7 months ago
I hit the same error today. It would be great if someone update the doc in get_started_with_triton_ensemble.ipynb
@katiemn can you please help update the notebook? Thanks.
Updated the notebook so the output returns successfully without a workaround
Expected Behavior
get_started_with_triton_ensemble.ipynb sample works without any modification, i.e.
get_triton_prediction_vertex
at CallingrawPredict
using Vertex AI SDK to get prediction response section returns the output .Actual Behavior
By following the above sample,
get_triton_prediction_vertex
fails with the following tracebackWorkaround
_get_inference_request
is under_utils.py
file now in triton-inference-server's client ref.custom_parameters
as is requiredBy applying the above changes, the request looks be sent without issue. The below is the updated function used for this purpose.
Steps to Reproduce the Problem
/notebooks/community/vertex_endpoints/nvidia-triton/get_started_with_triton_ensemble.ipynb
Specifications