Search results

In-process Python API: Server.stop() hangs until exit_timeout and never invokes Python-backend finalize()

Status: Open.

#8755 In triton-inference-server/server;

· bilelomrani1 opened on Apr 22, 2026

Wrong input numbers/input spec for torch aoti export model

bugSomething isn't workingSomething isn't working

Status: Open.

#8754 In triton-inference-server/server;

· geoffreyQiu opened on Apr 22, 2026

Ensemble Model only marginally faster than separate single-model calls -- is this expected?

Status: Open.

Task

#8740 In triton-inference-server/server;

· corentin87 opened on Apr 15, 2026

Why Does KIND_MODEL Validation Forbid gpus Field when Backend API Supports device_id for All Instance Kinds

Status: Open.

Task

#8715 In triton-inference-server/server;

· dhanushSB96 opened on Mar 27, 2026

vllm-omni support

Status: Open.

Feature

#8714 In triton-inference-server/server;

· hasanur-avatarin opened on Mar 25, 2026

speculative config with vllm backend

Status: Open.

Task

#8699 In triton-inference-server/server;

· geraldstanje opened on Mar 13, 2026

I got hold of 2 x Asus Ascent DGX and tested vLLM and Ollama

Status: Open.

#8691 In triton-inference-server/server;

· CG-8663 opened on Mar 10, 2026

Add Part 9: GPU-Accelerated Semantic Caching with cuVS CAGRA#149

Documentation

Status: Open.

Feature

#8677 In triton-inference-server/server;

· linear opened on Feb 25, 2026

CUDA IPC Handle failure in WSL2 forces CPU memory fallback, causing massive PCIe bottlenecks in BLS pipelines

bugSomething isn't workingSomething isn't working

EnhancementNew feature or requestNew feature or request

Status: Open.

Bug

#8670 In triton-inference-server/server;

· CarlosNacher opened on Feb 21, 2026

Provided serve file in container should allow override of model repository

EnhancementNew feature or requestNew feature or request

Status: Open.

Feature

#8668 In triton-inference-server/server;

· Pwalne opened on Feb 20, 2026

Segmentation fault on model reload when using Python backend metrics due to shared Metric object across processes

bugSomething isn't workingSomething isn't working

metricsRelated to deployment, inference, GPU metrics (prometheus, grafana, dcgm, etc.)Related to deployment, inference, GPU metrics (prometheus, grafana, dcgm, etc.)

Status: Open.

Bug

#8663 In triton-inference-server/server;

· aleksn7 opened on Feb 16, 2026

Segmentation fault on shutdown when using Python backend metrics

bugSomething isn't workingSomething isn't working

metricsRelated to deployment, inference, GPU metrics (prometheus, grafana, dcgm, etc.)Related to deployment, inference, GPU metrics (prometheus, grafana, dcgm, etc.)

Status: Open.

Bug

#8662 In triton-inference-server/server;

· aleksn7 opened on Feb 16, 2026