Search results

fix: MaxPodHitCount metric not registered in Collectors(), invisible to Prometheus

Status: Open.

#508 In llm-d/llm-d-kv-cache;

· wenhug opened on Apr 10, 2026

Broken tokenization for tool calls using --tool-call-parser qwen3_xml

Status: Open.

#506 In llm-d/llm-d-kv-cache;

· shashwatj07 opened on Apr 10, 2026

llmd_fs_connector: background write threads access freed GPU tensors under high concurrency

Status: Open.

#504 In llm-d/llm-d-kv-cache;

· natoscott opened on Apr 9, 2026

[Bug] Indexer blockSize not propagated from scheduler TokenProcessorConfig

bugSomething isn't workingSomething isn't working

Status: Open.

#490 In llm-d/llm-d-kv-cache;

· sagearc opened on Mar 31, 2026

ci: Add benchstat-based performance regression detection

Status: Open.

#489 In llm-d/llm-d-kv-cache;

· gyliu513 opened on Mar 31, 2026

fix: RenderChat gRPC timeout too short for multimodal requests

Status: Open.

#482 In llm-d/llm-d-kv-cache;

· vMaroon opened on Mar 30, 2026

Add HMA support to FS connector via vllm offloading connector HMA support

Status: Open.

#472 In llm-d/llm-d-kv-cache;

· kfirtoledo opened on Mar 29, 2026

Hybrid KV cache support for mamba+attention models (Qwen3.5)

Status: Open.

#465 In llm-d/llm-d-kv-cache;

· malaiwah opened on Mar 26, 2026

Strong-type tools and chat_template_kwargs in RenderChatRequest

Status: Open.

#463 In llm-d/llm-d-kv-cache;

· sagearc opened on Mar 25, 2026

Remove offsets from Tokenizer interface

Status: Open.

#462 In llm-d/llm-d-kv-cache;

· sagearc opened on Mar 25, 2026

Share same lint version for kv cache and GIE

Status: Open.

#459 In llm-d/llm-d-kv-cache;

· gyliu513 opened on Mar 25, 2026

llmd fs backend: EngineCore deadlock under SharedStorageOffloadingSpec + mp executor at high concurrency

Status: Open.

#457 In llm-d/llm-d-kv-cache;

· natoscott opened on Mar 24, 2026