Search results

Config Explorer should consider <code>--dtype</code> and <code>--kv-cache-dtype</code>

help wantedExtra attention is neededExtra attention is needed

Status: Open.

Feature

#420 In llm-d/llm-d-benchmark;

· jgchn opened on Oct 7, 2025

Parallelize CI/CD tests

help wantedExtra attention is neededExtra attention is needed

Status: Open.

Feature

#317 In llm-d/llm-d-benchmark;

· maugustosilva opened on Aug 29, 2025

Ensure that both <code>istio</code> and <code>kgateway</code> are tested

help wantedExtra attention is neededExtra attention is needed

Status: Open.

Feature

#316 In llm-d/llm-d-benchmark;

· maugustosilva opened on Aug 29, 2025

[Technical debt] Ensure the <code>python</code> implementation of <code>llmdbench_execute_cmd</code> has feature parity with the <code>bash</code> version.

help wantedExtra attention is neededExtra attention is needed

Status: Open.

Feature

#312 In llm-d/llm-d-benchmark;

· maugustosilva opened on Aug 28, 2025

Add universal report conversion to "deployment parameters" input

help wantedExtra attention is neededExtra attention is needed

Status: Open.

Feature

#219 In llm-d/llm-d-benchmark;

· maugustosilva opened on Aug 6, 2025

Running duplicate benchmark returns data from prior benchmarks

help wantedExtra attention is neededExtra attention is needed

Status: Open.

Bug

#157 In llm-d/llm-d-benchmark;

· namasl opened on Jul 21, 2025

"Non-standard" model names are not supported.

help wantedExtra attention is neededExtra attention is needed

Status: Open.

Bug

#156 In llm-d/llm-d-benchmark;

· namasl opened on Jul 21, 2025

Check for errors when running benchmarks

help wantedExtra attention is neededExtra attention is needed

Status: Open.

Feature

#96 In llm-d/llm-d-benchmark;

· namasl opened on Jul 2, 2025

Couple wait time for benchmarking run to complete to workload profile

help wantedExtra attention is neededExtra attention is needed

Status: Open.

Feature

#84 In llm-d/llm-d-benchmark;

· namasl opened on Jun 20, 2025