Chang GuoandCursor Agent 541128bdeb [KV Offload] Enable single-copy MLA layout for CPUOffloadingSpec (#50301) ...

Signed-off-by: Change72 <changg@nvidia.com>
Co-authored-by: Cursor Agent <cursor-agent@cursor.com>

2026-07-31 06:20:47 +03:00

..

assets

[Bugfix] Align OpenCV video metadata timeline (#47099)

2026-06-30 20:43:42 +00:00

benchmarks

[Bugfix][Benchmarks] Restore --skip-tokenizer-init with custom dataset (#49180)

2026-07-24 04:30:19 -07:00

compilation

[compile] Fix fake kernel return dtype (#50444)

2026-07-30 19:50:02 +00:00

config

[Model Runner V2][Spec Decode] Add multi-layer MTP speculator (#48892)

2026-07-30 15:54:15 -07:00

cute_utils

Revert "[Perf][GLM-5.2] Blackwell decode optimizations" (#49768)

2026-07-24 16:04:59 -07:00

device_allocator

[Bugfix][CuMem] Make KV-cache wake cleanup tag-safe (#49857)

2026-07-26 12:09:13 -07:00

distributed

[rl] Stateful Trainer Send: IPC [2/N] (#48981)

2026-07-30 21:59:08 +08:00

engine

[New model] Kimi K3 (#50000)

2026-07-30 03:49:58 -07:00

entrypoints

[PARSER][Mistral] unified engine-based parser for reasoning and tool calls (#48947)

2026-07-30 15:38:00 +02:00

inputs

[Frontend][Core] Standardize request error handling with VLLMError hierarchy (#49665)

2026-07-29 04:20:40 +00:00

ir

[CI] Use explicit devices in IR tests (#49513)

2026-07-23 20:08:14 -07:00

kernels

[Kernel][Helion] Disable unsafe B200 RMS reduction warp specialization (#50345)

2026-07-30 20:37:52 +00:00

logging_utils

…

lora

[Model] Add Inkling LoRA support [4/N] (#48884)

2026-07-17 09:52:15 +08:00

model_executor

[XPU] Fix FP8 block scale layout for MLA compatibility (#50349)

2026-07-31 11:05:35 +08:00

models

[Model Runner V2][Spec Decode] Add multi-layer MTP speculator (#48892)

2026-07-30 15:54:15 -07:00

multimodal

[Bugfix][Multimodal] Include media IO config in MM cache hash (#49975)

2026-07-29 10:48:40 +00:00

parser

[Frontend] Preserve bare Inkling text in Python and Rust parsers (#50403)

2026-07-30 12:25:10 -07:00

platforms

[XPU] [BugFix] Add deepseek_v4_fp8 to xpu supported_quantization list (#50434)

2026-07-31 11:12:04 +08:00

plugins

[Frontend] Add endpoint plugins framework (#47454)

2026-07-08 10:00:41 +08:00

profiler

…

ray

Propagate Flash Attention cache configuration to Ray workers (#49177)

2026-07-21 07:47:53 +00:00

reasoning

[PARSER][Mistral] unified engine-based parser for reasoning and tool calls (#48947)

2026-07-30 15:38:00 +02:00

renderers

[Frontend] Add detokenization streaming derender for disaggregated serving (#47301)

2026-07-29 18:01:20 -07:00

third_party

[New model] Kimi K3 (#50000)

2026-07-30 03:49:58 -07:00

tokenizers

[PARSER][Mistral] unified engine-based parser for reasoning and tool calls (#48947)

2026-07-30 15:38:00 +02:00

tool_parsers

[Frontend][Bugfix] Use default tool call IDs for Kimi K3 for conversation-level uniqueness (#50420)

2026-07-31 10:02:17 +08:00

tracing

…

transformers_utils

[New model] Kimi K3 (#50000)

2026-07-30 03:49:58 -07:00

triton_utils

[ROCm] [Model] Enable TML inkling (#48841)

2026-07-27 10:05:46 +00:00

usage

…

utils

Enable gfx1250 ROCm architecture (#46516)

2026-07-31 01:41:12 +00:00

v1

[KV Offload] Enable single-copy MLA layout for CPUOffloadingSpec (#50301)

2026-07-31 06:20:47 +03:00

vllm_flash_attn

[CI/Build][Docker] Bump nvidia-cutlass-dsl to 4.6.0 and drop packaging workarounds (#47442)

2026-07-16 13:51:41 +00:00

__init__.py

…

_aiter_ops.py

Enable gfx1250 ROCm architecture (#46516)

2026-07-31 01:41:12 +00:00

_custom_ops.py

[CPU] Bump up CPU kernels to latest version (#50387)

2026-07-30 19:18:43 +08:00

_xpu_ops.py

[XPU][DeepSeekV4]Add DeepSeek-V4 fuse_index_q SYCL kernel path (#45991)

2026-07-21 21:22:34 +08:00

collect_env.py

…

connections.py

…

env_override.py

[Perf] Add VLLM_TRITON_FORCE_FIRST_CONFIG to skip Triton autotuning (#42425)

2026-06-16 15:16:45 +02:00

envs.py

[New model] Kimi K3 (#50000)

2026-07-30 03:49:58 -07:00

exceptions.py

[Frontend][Core] Standardize request error handling with VLLMError hierarchy (#49665)

2026-07-29 04:20:40 +00:00

forward_context.py

[Feature] Support sequence parallel without the need for DP, 1.9%~5.0% E2E Throughput Improvement (#47070)

2026-07-05 05:41:30 -07:00

logger.py

…

logits_process.py

…

logprobs.py

…

model_inspection.py

…

outputs.py

[Front-end] [Messages] Populate num_cache_creation_tokens (#48535)

2026-07-18 13:04:35 -04:00

pooling_params.py

[Frontend][Core] Standardize request error handling with VLLMError hierarchy (#49665)

2026-07-29 04:20:40 +00:00

py.typed

…

sampling_params.py

[Frontend][Core] Standardize request error handling with VLLMError hierarchy (#49665)

2026-07-29 04:20:40 +00:00

scalar_type.py

Humming support for 2/3/5/6/7-bit pack-quantized weight-only inference (#46389)

2026-06-24 13:53:54 -04:00

scripts.py

…

sequence.py

[Core] Cleanup KVConnector handling with PP + fix MRV2 (#43732)

2026-05-28 13:12:03 -07:00

tasks.py

[Bugfix] Reject removed pooling parameters (#48984)

2026-07-19 05:18:03 -07:00

version.py

…