3 people e83222396f [ROCm][Perf] W4A16: speed up skinny GEMM decode, including speculative batches (#57199) ...

Signed-off-by: Robert Esclapez Garcia <robert.garcia@amd.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Co-authored-by: Douglas Lehr <91553416+dllehr-amd@users.noreply.github.com>

2026-10-13 17:21:30 +00:00

..

basic_correctness

[MRV2] Switch CI tests still running on MRV1 unnecessarily (#61014)

2026-10-11 09:57:07 -07:00

benchmarks

[Frontend] Support min_p in the Responses API (#48084)

2026-10-08 20:38:17 +00:00

compile

[ROCm][Perf] Fix Gemma AR+RMS prefill fallback (#58547)

2026-10-12 19:53:08 +00:00

config

[Model] Add Clef-Flash support (#61308)

2026-10-13 05:46:37 +00:00

cuda

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

detokenizer

[Bugfix][Spec Decode] Trim token_ids/logprobs left past a stop string under speculative decoding (#47616)

2026-10-09 14:22:31 -07:00

device_allocator

[Bugfix] Don't drop the rest of the allocator config when toggling expandable segments (#57982)

2026-09-25 17:34:19 -05:00

distributed

[ROCm][Bugfix] Run the AITER AR+RMSNorm fallback test on the platform… (#61350)

2026-10-13 08:35:57 +00:00

engine

[Perf] Support minimum KV splits configuration for TokenSpeed MLA decode (#60345)

2026-10-09 23:33:47 +00:00

entrypoints

[Misc] Break chat_utils into more maintainable modules. (#60972)

2026-10-13 15:08:07 +00:00

evals

[Core][Mamba] Remove leftovers of mamba_cache_mode="all" (#60043)

2026-10-13 00:58:59 -07:00

fusion

[Fusion] Support 3D activation quantization fusion and enable it for Pixtral (#59612)

2026-10-09 15:12:03 +08:00

ir

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

jit_monitor

[Perf] JIT monitor: hook TileLang JITImpl.compile instead of __call__ (#58685)

2026-10-08 00:11:59 +00:00

kernels

[ROCm][Perf] W4A16: speed up skinny GEMM decode, including speculative batches (#57199)

2026-10-13 17:21:30 +00:00

lora

[Bugfix][LoRA] Apply PEFT rank_pattern / alpha_pattern scaling per module (#59801)

2026-10-12 21:49:30 +08:00

model_executor

[Quantization] Add ModelOpt IQ2_XS, IQ2_XXS, and Q8_0 support via b12x (#59604)

2026-10-13 09:06:33 -07:00

models

[Misc] Break chat_utils into more maintainable modules. (#60972)

2026-10-13 15:08:07 +00:00

multimodal

[Security] Bound decoded and downloaded media per request (#61226)

2026-10-13 10:17:14 +00:00

parser

[Bugfix] Accept whole-number floats for integer tool arguments in the response parser (#61297)

2026-10-13 12:55:20 +08:00

platforms

[Attention] Remove cascade attention (#61126)

2026-10-12 10:41:07 -07:00

plugins

[Bugfix][Platform] Don't resolve current_platform during import vllm (#60766)

2026-10-12 17:05:34 +00:00

plugins_tests

[Bugfix][Platform] Don't resolve current_platform during import vllm (#60766)

2026-10-12 17:05:34 +00:00

prompts

…

quantization

[Quantization] Add ModelOpt IQ2_XS, IQ2_XXS, and Q8_0 support via b12x (#59604)

2026-10-13 09:06:33 -07:00

reasoning

[Frontend] Port Step-3.5 parsers to the streaming parser engine (#59321)

2026-10-01 22:18:37 +00:00

renderers

[Misc] Break chat_utils into more maintainable modules. (#60972)

2026-10-13 15:08:07 +00:00

rocm

[ROCm][CI] Add test coverage for VLLM_ROCM_MOE_PADDING memory-stride padding transparency (#59332)

2026-10-02 11:23:22 -05:00

samplers

[watermark] compatibility validation (#56801)

2026-10-05 14:37:13 -04:00

spec_decode

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

standalone_tests

[CI] Add nightly Python 3.14 smoke test (#60456)

2026-10-12 10:11:26 -07:00

tokenizers_

[Bugfix] Preserve independent tokenizer code_revision in snapshot resolution (#61272)

2026-10-13 04:23:45 +00:00

tool_parsers

[Frontend] Support CustomToolCall in the non-Harmony Responses API (#57322)

2026-10-12 17:22:55 +08:00

tool_use

[Frontend] Support CustomToolCall in the non-Harmony Responses API (#57322)

2026-10-12 17:22:55 +08:00

tools

[Agents] Add pre-commit check for agent files and skills (#59657)

2026-10-02 09:38:22 -07:00

tracing

[Refactor] Remove dead tests code (#58916)

2026-09-30 04:39:03 +00:00

transformers_utils

[Performance] Load HF tokenizers and processors from a local snapshot (#60825)

2026-10-12 13:45:54 +00:00

utils_

[Bugfix] Scope deprecated CLI argument warnings to the parser that defines them (#59917)

2026-10-08 15:29:04 +00:00

v1

[Bugfix] Don't split a multimodal item that a prefill chunk starts exactly on (#52412)

2026-10-13 15:32:38 +00:00

vllm_test_utils

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

watermarking

[Metrics] Count finished requests by watermark status (#58715)

2026-10-12 17:21:22 +00:00

__init__.py

…

ci_envs.py

Add pydocstyle to the ruff rules (#52136)

2026-09-16 16:44:29 +00:00

conftest.py

[Attention] Remove cascade attention (#61126)

2026-10-12 10:41:07 -07:00

test_attention_backend_registry.py

[Platform] Allow BackendEnum dynamic register (#41091)

2026-10-08 09:56:19 +08:00

test_audio_media_size_precheck.py

[Security] Bound decoded and downloaded media per request (#61226)

2026-10-13 10:17:14 +00:00

test_cmake_utils.py

[CI] Ignore ruff D209, rejoin the docstrings it split, and silence incompatible-rule warnings (#57212)

2026-09-16 10:57:24 -07:00

test_config.py

[Qwen4Exp] support Sequence Parallelism for Qwen3.8-Flash-Next (#56322)

2026-10-12 20:56:44 +08:00

test_connections.py

[Security] Bound decoded and downloaded media per request (#61226)

2026-10-13 10:17:14 +00:00

test_embedded_commit.py

…

test_envs.py

[Security] Fix P2P ZMQ socket exhaustion via unbounded peer sessions (#51504)

2026-10-11 11:39:11 -04:00

test_force_first_config.py

…

test_fxgraphcache_pickle_patch.py

…

test_logger.py

[Core][Logging] Add built-in JSON formatter (#58739)

2026-10-01 09:11:46 +01:00

test_logprobs.py

[Bugfix] Handle empty FlatLogprobs slices and delta output (#53704)

2026-08-26 21:00:54 +08:00

test_outputs.py

[Bugfix][Core] Preserve speculative metrics when coalescing outputs (#55708)

2026-10-07 08:35:14 +08:00

test_pcp_dp.py

[PCP] Support prefill context parallelism with data parallelism (#57075)

2026-09-23 22:32:36 -04:00

test_pooling_params.py

[Feature][Pooling] Add offline fixed-length late chunking (#60164)

2026-10-13 11:16:21 +00:00

test_quantization_revision_pin.py

Preserve revision pins in secondary artifact loaders (#51446)

2026-08-11 04:33:10 +00:00

test_ray_env_utils.py

[Bugfix] Set breakable graph env before Ray actor import (#53293)

2026-08-28 23:11:50 +00:00

test_ray_env.py

[CI] Ignore ruff D209, rejoin the docstrings it split, and silence incompatible-rule warnings (#57212)

2026-09-16 10:57:24 -07:00

test_regression.py

…

test_request_input_bounds.py

[CI] Allowlist-shrink batch 1: wire 17 root-level tests + drop 5 stale watermarking entries into misc.yaml (#58055)

2026-09-27 22:23:46 -07:00

test_sampling_params.py

[watermark] compatibility validation (#56801)

2026-10-05 14:37:13 -04:00

test_scalartype.py

…

test_sequence.py

…

test_tensorizer_revision_pin.py

Preserve revision pins in secondary artifact loaders (#51446)

2026-08-11 04:33:10 +00:00

test_triton_utils.py

[CI] Allowlist-shrink batch 1: wire 17 root-level tests + drop 5 stale watermarking entries into misc.yaml (#58055)

2026-09-27 22:23:46 -07:00

test_uvicorn_logging.py

[Core][Logging] Add built-in JSON formatter (#58739)

2026-10-01 09:11:46 +01:00

test_version.py

[CI] Allowlist-shrink batch 1: wire 17 root-level tests + drop 5 stale watermarking entries into misc.yaml (#58055)

2026-09-27 22:23:46 -07:00

test_vllm_port.py

…

test_zen_cpu_platform_detection.py

[CI] Allowlist-shrink batch 2: wire 9 tests/models/ files into Buildkite, fix 2 stale bugs found (#59050)

2026-10-06 21:16:53 -07:00

test_zentorch_version_pin.py

[ZenCPU][Test] Add zentorch/torch version pin consistency test (#52888)

2026-10-06 13:12:19 +08:00

utils.py

[MRV2] Switch CI tests still running on MRV1 unnecessarily (#61014)

2026-10-11 09:57:07 -07:00