Ben Browning

Consider formatting code with `gofmt`

5

Many Golang project use `gofmt` to consistently format their Go code. And, it's not uncommon to setup your editor to automatically run `gofmt` when saving changes to a file. Perhaps...

Knative Service invocation should use status.url instead of Istio IP

3

Knative Services should be invoked via the Service's `status.url` property instead of setting a host header and directly trying to hit the Istio ingress gateway. Not all Knative installations use...

### Your current environment Running inside an OpenShift cluster, but not directly relevant for the issue in question. ### 🐛 Describe the bug At https://github.com/vllm-project/vllm/blob/2d7bce9cd5981db146b18a8a95c5a7e0480687bd/vllm/model_executor/model_loader/tensorizer.py#L74-L80 the tensorizer loader checks if...

Worker volume name hardcoded and may be wrong in crc_libvirt.sh

5

On line 116 and the 2nd argument on line 118 of https://github.com/praveenkumar/osp4/blob/6f7d3ea9bf2ac194aceac4b3ae1be81d509c6615/libvirt/crc_libvirt.sh#L116 you see the actual on-disk filename of the worker volume is hardcoded. However, when building, the last part...

Don't assume default storage pool is at /var/lib/libvirt/images in crc_libvirt.sh and test1-worker-0-98nsr.xml

1

This script checks that the default storage pool exists and if it doesn't, it creates it at /var/lib/libvirt/images. However, later on in the script, it assumes that the default pool...

# What does this PR do? This adds some initial content documenting our OpenAI compatible APIs - Responses, Chat Completions, Completions, and Models - along with instructions on how to...

### System Info Latest llama-stack from main commit 2603f10f95fcd302297158adb709d2a84c9f60af ### 🐛 Describe the bug When testing the responses API with the ollama provider, I'm getting frequent errors with unique id...

Multiple providers blocking the async event loop

11

### 🐛 Describe the bug Llama Stack uses FastAPI and an async event loop. FastAPI uses a single event loop to dispatch requests to all async request handlers. If this...

Multiple inference providers unconditionally adding BOS token to completion prompts

2

### System Info Latest llama-stack from main as of this bug report (029e4fc64d9017eed625c927a69e71fff9033727) ### 🐛 Describe the bug When calling the completions API, multiple inference providers are adding Llama-specific BOS...

Responses API: Delete a response API endpoint

2

### 🚀 Describe the new functionality needed We're missing the API endpoint to delete Responses from the conversation state - see https://platform.openai.com/docs/api-reference/responses/delete . ### 💡 Why is this needed? What...