-
Notifications
You must be signed in to change notification settings - Fork 7
Insights: neuralmagic/gateway-api-inference-extension
Overview
Could not load contribution data
Please try again later
24 Pull requests merged by 9 people
-
Implemented KVCacheAwareScorer
#34 merged
Apr 30, 2025 -
[fix]: Add oc check to kubernetes-dev-env script
#90 merged
Apr 30, 2025 -
Minor fixes in instructions for deploying dev k8s env
#58 merged
Apr 30, 2025 -
Sessions support is removed from datastore, server and scheduler
#88 merged
Apr 30, 2025 -
pd filter returns full url of the port
#94 merged
Apr 30, 2025 -
test: Test load based scorer
#81 merged
Apr 30, 2025 -
update prefill header name
#89 merged
Apr 29, 2025 -
Pd filter
#87 merged
Apr 29, 2025 -
Read pod role from kubernetes pod info in reconciler
#82 merged
Apr 29, 2025 -
feat: Add scripts for kubernetes dev env using vLLM and vLLM-p2p
#60 merged
Apr 29, 2025 -
Add local_config.go which adds custom plugins
#86 merged
Apr 29, 2025 -
Dev rebase 2025 04 29
#84 merged
Apr 29, 2025 -
Add load based scorer
#77 merged
Apr 29, 2025 -
Rebase main into dev - 2025-04-28
#74 merged
Apr 28, 2025 -
Rebased latest K8S Main into dev
#61 merged
Apr 27, 2025 -
Report lint errors on new code only
#54 merged
Apr 27, 2025 -
feat: add cleanup for kubernetes dev env
#59 merged
Apr 24, 2025 -
Add max score picker
#51 merged
Apr 24, 2025 -
Makefile target setting prod and dev versions to the same value
#55 merged
Apr 24, 2025 -
fix: Correct the setup of the KGateway Envoy service to be a proper NodePort service
#57 merged
Apr 24, 2025 -
tweak instructions for accessing inference service on kind cluster
#56 merged
Apr 24, 2025 -
test: Scorers test with dummy scorer and picker
#53 merged
Apr 24, 2025 -
Add Kubernetes dev environment deployments with KGateway
#50 merged
Apr 24, 2025
4 Pull requests opened by 3 people
-
Session affinity scorer
#52 opened
Apr 24, 2025 -
Partial implementation of Session Affinity (without headers in response)
#97 opened
Apr 30, 2025 -
Added unit test for pod role loading
#99 opened
Apr 30, 2025 -
Change quay.io path
#102 opened
Apr 30, 2025
11 Issues closed by 6 people
-
Change prefill header value to URL
#92 closed
Apr 30, 2025 -
Remove sessions support from datastore, server and scheduler
#95 closed
Apr 30, 2025 -
Test load based scorer
#80 closed
Apr 30, 2025 -
Read pod's role (p/d/both) from kubernetes label
#78 closed
Apr 30, 2025 -
Add initial version of p/d filter
#83 closed
Apr 30, 2025 -
Create local_config.go which will contain all specific plugins
#85 closed
Apr 30, 2025 -
Development Environments
#3 closed
Apr 29, 2025 -
Deployments - Cleanup & Documentation
#28 closed
Apr 29, 2025 -
OpenShift Dev Environment - Full Gateway+GIE Stack Deployment with VLLM and VLLM-P2P mode
#72 closed
Apr 29, 2025 -
Endpoint Picker logger not showing traces
#45 closed
Apr 29, 2025 -
Add load based routing support
#66 closed
Apr 29, 2025
17 Issues opened by 8 people
-
Migrate to llm-d image repo
#101 opened
Apr 30, 2025 -
Required env vars not blocked in kubernetes dev env
#100 opened
Apr 30, 2025 -
Add unit test for pod role loading
#98 opened
Apr 30, 2025 -
Create sesssion affinity scorer
#96 opened
Apr 30, 2025 -
EPP should be able to set full pod url in header for P/D purposes
#93 opened
Apr 30, 2025 -
Add unit tests for p-d filter
#91 opened
Apr 30, 2025 -
Adding Prometheus and Grafana for vLLM Monitoring in the Deployment Setup
#75 opened
Apr 28, 2025 -
Scorers/Filters needs access (read/update) to Request/Response Headers/Body
#73 opened
Apr 28, 2025 -
Support for KV cache aware routing
#71 opened
Apr 27, 2025 -
Incorporate post-build tests in CICD pipeline
#70 opened
Apr 27, 2025 -
Add support for disaggregated Prefill/Decode (PD)
#69 opened
Apr 27, 2025 -
Support multiple instances of the EPP for HA/load
#68 opened
Apr 27, 2025 -
Add internal queueing of requests
#67 opened
Apr 27, 2025 -
Explore reuse of GIE reference implementation to support proposed filter/scorer paradigm
#65 opened
Apr 27, 2025 -
Document process for deploying router EPP
#64 opened
Apr 27, 2025 -
Use Kubernetes controller runtime
#63 opened
Apr 27, 2025 -
Run K8s GIE benchmark on a regular basis
#62 opened
Apr 27, 2025
5 Unresolved conversations
Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.
-
Prefix Aware Scorer
#48 commented on
Apr 30, 2025 • 16 new comments -
Add scorer based on active loras
#43 commented on
Apr 27, 2025 • 4 new comments -
Prefix Aware Routing
#42 commented on
Apr 27, 2025 • 0 new comments -
Fix in scorer manager in picking the best target
#35 commented on
Apr 27, 2025 • 0 new comments -
Add second vLLM-sim instance to allow for routing
#41 commented on
Apr 27, 2025 • 0 new comments