All issues

Search results

Remove deprecated WVA support from KServe

Status: Open.

#6262 In kserve/kserve;

· tessapham opened on Sep 21, 2026

Support InferenceService environment variables for storage-initializer authentication

good first issueGood for newcomersGood for newcomers

Status: Open.

#6260 In kserve/kserve;

· andresllh opened on Sep 21, 2026

[Bug]: huggingfaceserver liveness probe returns 200 OK after underlying vLLM engine crashes

Status: Open.

#6259 In kserve/kserve;

· zoramt opened on Sep 21, 2026

[SGLang Runtime] Follow-up: address gaps in KServe + SGLang integration

Status: Open.

#6250 In kserve/kserve;

· haitwang-cloud opened on Sep 21, 2026

HPA autoscaler silently ignores stabilizationWindowSeconds / autoscaling behavior configuration

Status: Open.

#6246 In kserve/kserve;

· zhouyou9505 opened on Sep 20, 2026

Incorrect Prometheus metrics port in kserve-mlserver runtime (8080 instead of 8082)

Status: Open.

#6227 In kserve/kserve;

· xrwang8 opened on Sep 17, 2026

Implement MCV capture and OCI packaging for GPU kernel caches

Status: Open.

#6223 In kserve/kserve;

· Jooho opened on Sep 17, 2026

Add v1alpha1 APIs for GPU kernel cache management

Status: Open.

#6220 In kserve/kserve;

· Jooho opened on Sep 16, 2026

The ghcr.io/llm-d/llm-d-inference-scheduler:v0.7.1 image can not be pulled

Status: Open.

#6215 In kserve/kserve;

· umka1332 opened on Sep 15, 2026

v2 inference returns 500 for models whose outputs are numpy unicode arrays

Status: Open.

#6204 In kserve/kserve;

· rishabhsai opened on Sep 14, 2026

from_np_dtype crashes with AttributeError on numpy scalar types and Python built-in types

Status: Open.

#6202 In kserve/kserve;

· Madhur-7568 opened on Sep 13, 2026

Tracking: inference logger reliability & observability fixes

Status: Open.

#6193 In kserve/kserve;

· ncapilla opened on Sep 11, 2026