Search results

[Perf Regression] 17 config(s) regressed @ 117f7d2c

performance

regression

Status: Open.

#403 In ROCm/ATOM;

· github-actions opened on Mar 25, 2026

[Perf Regression] 30 config(s) regressed @ f2213328

performance

regression

Status: Open.

#392 In ROCm/ATOM;

· github-actions opened on Mar 24, 2026

ATOM fails to load DeepSeek-V2-Lite-Chat and Qwen1.5-MoE-A2.7B-Chat (MLA/MoE models)

Status: Open.

#381 In ROCm/ATOM;

· sunway513 opened on Mar 23, 2026

[Perf Regression] 24 config(s) regressed @ 613e57af

performance

regression

Status: Open.

#380 In ROCm/ATOM;

· github-actions opened on Mar 23, 2026

[Feature]: how do i use disagg serving in ATOM?

Status: Open.

#377 In ROCm/ATOM;

· functionstackx opened on Mar 22, 2026

[RFC] Introduce ATOM as out of tree backend of SGLang

Status: Open.

#359 In ROCm/ATOM;

· zhuyuhua-v opened on Mar 19, 2026

[RFC] Introduce ATOM as model implementation backend of SGLang

Status: Open.

#295 In ROCm/ATOM;

· zejunchen-zejun opened on Mar 10, 2026

CI/CD: Add incremental test coverage for Plugin Mode (vLLM/SGLang)

Status: Open.

#255 In ROCm/ATOM;

· sunway513 opened on Mar 3, 2026

[Issue]: ValueError: could not broadcast input array from shape (58713,) into shape (117001,) with prefix cache enabled

Status: Open.

#238 In ROCm/ATOM;

· peizhang56 opened on Feb 25, 2026

[RFC] CK-Free AITER Compatibility: Attention, MOE, and Docker Changes

Status: Open.

#229 In ROCm/ATOM;

· sunway513 opened on Feb 23, 2026

[Issue]: ATOM fails on Qwen3 model when the flag "--enable_prefix_caching" is enabled

Status: Open.

#221 In ROCm/ATOM;

· vecheruk-amd opened on Feb 16, 2026

[RFC] Enable ATOM as vLLM out-of-tree Platform

Status: Open.

#201 In ROCm/ATOM;

· zejunchen-zejun opened on Feb 9, 2026