Fix ATOM image checkout for target commits

eb0ec84

Select commit

Loading

Failed to load commit list.

Uh oh!

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

Checks on: pull_request 2

Check Repository Dependency

Check Code Style with Black

Check Code Style with Ruff

Aiter Test on: pull_request 6

check-signal

build_aiter_wheels

prepare_triton_wheel

split_aiter_tests

Multi-GPU Tests (8 GPU)

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 5, 0)

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

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 5, 2)

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 5, 3)

Standard Tests (1 GPU) (linux-aiter-mi35x-1, MI35X, 5, 4)

Standard Tests (1 GPU) (aiter-1gpu-runner, MI300X, 5, 0)

Standard Tests (1 GPU) (aiter-1gpu-runner, MI300X, 5, 1)

Standard Tests (1 GPU) (aiter-1gpu-runner, MI300X, 5, 2)

Standard Tests (1 GPU) (aiter-1gpu-runner, MI300X, 5, 3)

Standard Tests (1 GPU) (aiter-1gpu-runner, MI300X, 5, 4)

Standard Test Results

Tuned Op Bench

Sglang Downstream Test on: pull_request 2

check-signal

select-sglang-tests

Sglang DeepSeek-R1-MXFP4 Accuracy (8 GPU)

Sglang Qwen 3.5 Accuracy (8 GPU)

Sglang DeepSeek-V3.2 Accuracy (8 GPU)

Triton Test on: pull_request 12

check-signal

split_triton_tests

prepare_triton_wheel

Triton Tests (MI35X) / Shard 0

Triton Tests (MI35X) / Shard 1

Triton Tests (MI35X) / Shard 2

Triton Tests (MI35X) / Shard 3

Triton Tests (MI35X) / Shard 4

Triton Tests (MI35X) / Shard 5

Triton Tests (MI35X) / Shard 6

Triton Tests (MI35X) / Shard 7

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

Triton Test Results

Triton MI300X Test Results

vLLM Benchmark on: pull_request 5

check-signal

Build Aiter wheel

vLLM Benchmark (8 GPU) (openai/gpt-oss-120b, fp8_kvcache, 8, --gpu-memory-utilization 0.3)

vLLM Benchmark (8 GPU) (deepseek-ai/DeepSeek-R1-0528, default_kvcache, 8, --block-size 1)

vLLM Benchmark (8 GPU) (moonshotai/Kimi-K2.5, default_kvcache, 8, --block-size 1 --trust-remote-c...

Flash Attention Integration on: pull_request 10

check-signal

prechecks

Flash Attention - Triton / MI35X (1 GPU)

Flash Attention - Triton / RDNA3 (1 GPU)

Windows Smoke Test

Flash Attention - CK (1 GPU)

OPUS Test on: pull_request 3

check-signal

OPUS Tests (MI35X)

OPUS Tests (MI300X)

Atom Test on: pull_request 16

check-signal

load-atom-models

Accuracy (Meta-Llama-3-8B-Instruct, meta-llama/Meta-Llama-3-8B-Instruct, --kv_cache_dtype fp8 --gpu-memory-utilization 0.3, linux-aiter-mi35x-1)

Accuracy (DeepSeek-R1-0528, deepseek-ai/DeepSeek-R1-0528, --kv_cache_dtype fp8 -tp 8, linux-aiter-mi35x-8)

Accuracy (DeepSeek-V4-Pro, deepseek-ai/DeepSeek-V4-Pro, --kv_cache_dtype fp8 -tp 8, linux-aiter-mi35x-8)

Accuracy (DeepSeek-V4-Pro MTP, deepseek-ai/DeepSeek-V4-Pro, --kv_cache_dtype fp8 -tp 8 --method mtp --num-speculative-tokens 3, linux-aiter-mi35x-8)

Accuracy (DeepSeek-R1-0528 MTP, deepseek-ai/DeepSeek-R1-0528, --kv_cache_dtype fp8 -tp 8 --method mtp --num-speculative-tokens 3, linux-aiter-mi35x-8)

Accuracy (gpt-oss-120b, openai/gpt-oss-120b, --kv_cache_dtype fp8 --gpu-memory-utilization 0.5, linux-aiter-mi35x-1)

Accuracy (Llama-3.3-70B-Instruct-MXFP4-Preview, amd/Llama-3.3-70B-Instruct-MXFP4-Preview, --kv_cache_dtype fp8 --gpu-memory-utilization 0.3, linux-aiter-mi35x-1)

Accuracy (DeepSeek-R1-0528-FP4, amd/DeepSeek-R1-0528-MXFP4-MTP-MoEFP4, --kv_cache_dtype fp8 -tp 4, linux-aiter-mi35x-8)

Accuracy (DeepSeek-R1-0528-FP4 MTP, amd/DeepSeek-R1-0528-MXFP4-MTP-MoEFP4, --kv_cache_dtype fp8 -tp 4 --method mtp --num-speculative-tokens 3, linux-aiter-mi35x-8)

Accuracy (Qwen3-235B-A22B-Instruct-2507-FP8, Qwen/Qwen3-235B-A22B-Instruct-2507-FP8, --kv_cache_dtype fp8 -tp 8 --enable-expert-parallel, linux-aiter-mi35x-8)

Accuracy (Qwen3-Next-80B-A3B-Thinking, Qwen/Qwen3-Next-80B-A3B-Thinking, -tp 8 --kv_cache_dtype fp8, linux-aiter-mi35x-8)

Accuracy (gpt-oss-120b (2 GPUs), openai/gpt-oss-120b, --kv_cache_dtype fp8 -tp 2 --enable-dp-attention --enable-expert-parallel --enable-tbo all --gpu-memory-utilization 0.5, linux-aiter-mi35x-4)

Accuracy (Kimi-K2.5-MXFP4, amd/Kimi-K2.5-MXFP4, --kv_cache_dtype fp8 -tp 4 --trust-remote-code, linux-aiter-mi35x-8)

Accuracy (GLM-5.1-FP8, zai-org/GLM-5.1-FP8, --kv_cache_dtype fp8 -tp 8 --default-chat-template-kwargs '{"enable_thinking":false}', linux-aiter-mi35x-8)

Accuracy (GLM-5.1-MXFP4 MTP, amd/GLM-5.1-MXFP4, --kv_cache_dtype fp8 -tp 4 --default-chat-template-kwargs '{"enable_thinking":false}' --method mtp --num-speculative-tokens 3, linux-aiter-mi35x-8)

Accuracy (GLM-5.1-MXFP4, amd/GLM-5.1-MXFP4, --kv_cache_dtype fp8 -tp 4 --default-chat-template-kwargs '{"enable_thinking":false}', linux-aiter-mi35x-8)

Accuracy (Qwen3.5-397B-A17B-FP8 MTP, Qwen/Qwen3.5-397B-A17B-FP8, --kv_cache_dtype fp8 -tp 4 --method mtp --num-speculative-tokens 3, linux-aiter-mi35x-8)

Accuracy (Qwen3.5-397B-A17B-MXFP4, amd/Qwen3.5-397B-A17B-MXFP4, --kv_cache_dtype fp8 -tp 4, linux-aiter-mi35x-8)

Accuracy (Qwen3.5-397B-A17B-MXFP4 MTP, amd/Qwen3.5-397B-A17B-MXFP4, --kv_cache_dtype fp8 -tp 4 --method mtp --num-speculative-tokens 3, linux-aiter-mi35x-8)

Fix ATOM image checkout for target commits

Uh oh!

Select a check to view from the sidebar