Merge branch 'main' into feat/gfx942-fmha-hd192-split-asm

d0b16eb

Select commit

Loading

Failed to load commit list.

Uh oh!

There was an error while loading. Please reload this page.

Kimi Downstream Test on: pull_request

Kimi-K2.5 (${{ matrix.backend }})

Checks on: pull_request 5

Check Repository Dependency

Check Code Style with Black

Check Code Style with Ruff

PR Title Tags & Labels on: pull_request_target 3

tag-title

Extended Test on: pull_request_target 1

Dispatch extended tests

Sglang Downstream Test on: pull_request

ci_config

check-signal

select-sglang-tests

Sglang ${{ matrix.model }} ${{ matrix.test_type }} (8 GPU)

vLLM disagg CI smoke workflow on: pull_request

ci_config

Select upstream vLLM cases

Build gfx950 Aiter overlay wheels

matrix.case_name

OPUS Test on: pull_request 5

ci_config / config

check-signal

OPUS Tests (MI35X)

OPUS Tests (MI300X)

Flash Attention Integration on: pull_request 9

ci_config / config

check-signal

prechecks

Flash Attention - Triton / ${{ matrix.label }} (1 GPU)

Windows Smoke Test

Triton-only Smoke

Flash Attention - CK (1 GPU)

FA MHA Bench

Aiter Test on: pull_request 37

Detect Aiter Test Scope

ci_config / config

check-signal

build_aiter_wheels (3.10, rocm/pytorch:rocm7.2.4_ubuntu22.04_py3.10_pytorch_release_2.10.0, false)

build_aiter_wheels (3.12, rocm/pytorch:rocm7.2.4_ubuntu24.04_py3.12_pytorch_release_2.10.0, true)

prepare_triton_wheel / prepare

Upload wheels and manifest to S3

split_aiter_tests

Detect Multi-GPU Test Scope

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 8, 0)

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 8, 1)

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 8, 2)

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 8, 3)

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 8, 4)

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 8, 5)

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 8, 6)

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 8, 7)

Standard Tests (1 GPU) (linux-aiter-oci-mi300x-1, MI300X, 8, 0)

Standard Tests (1 GPU) (linux-aiter-oci-mi300x-1, MI300X, 8, 1)

Standard Tests (1 GPU) (linux-aiter-oci-mi300x-1, MI300X, 8, 2)

Standard Tests (1 GPU) (linux-aiter-oci-mi300x-1, MI300X, 8, 3)

Standard Tests (1 GPU) (linux-aiter-oci-mi300x-1, MI300X, 8, 4)

Standard Tests (1 GPU) (linux-aiter-oci-mi300x-1, MI300X, 8, 5)

Standard Tests (1 GPU) (linux-aiter-oci-mi300x-1, MI300X, 8, 6)

Standard Tests (1 GPU) (linux-aiter-oci-mi300x-1, MI300X, 8, 7)

Multi-GPU Tests (8 GPU)

Standard Test Results

Tuned Op Bench

Aiter Test Gate

vLLM Benchmark on: pull_request

ci_config

check-signal

Build Aiter wheel

vLLM Benchmark

Triton Test on: pull_request 3

split_triton_tests

ci_config

check-signal

prepare_triton_wheel

Triton Tests (MI35X) / Shard ${{ matrix.shard }}

Triton Tests (MI300X) / Shard ${{ matrix.shard }}

Triton Test Results

Atom Test on: pull_request

ci_config

check-signal

load-atom-models

Accuracy (${{ matrix.display_name }}, ${{ matrix.runner }})

Select a check to view from the sidebar