sglang

arg_groups

Custom spec algorithm can handle server args (#28162 )

2026-06-16 17:13:52 -07:00

batch_invariant_ops

[NPU]adaptation to support deterministic inference (#21197 )

2026-06-09 09:22:09 +08:00

batch_overlap

Add get_parallel(): a structured accessor for parallel-topology state (#28567 )

2026-06-17 20:23:43 -07:00

checkpoint_engine

update pre-commit config (#18860 )

2026-02-16 00:18:31 +08:00

compilation

fix(fa3): no NaN embeddings with fa_skip_kv_cache under piecewise CUDA graph (#27343 )

2026-06-15 13:46:22 -07:00

configs

Revert "revert the head_dim assignment from PR 23862" (#28583 )

2026-06-17 17:11:05 -07:00

connector

[connector] Add Azure Blob Storage connector (az:// and *.blob.core.windows.net URLs) (#23995 )

2026-05-11 14:20:22 -07:00

constrained

fix: prevent stale bitmask leakage in LLGuidance grammar backend (#23653 )

2026-06-13 13:27:08 -07:00

debug_utils

[lint] Enable Ruff UP037 to drop redundant quoted annotations (#27984 )

2026-06-11 17:38:05 -07:00

disaggregation

[PD] Optimize SWA allocation (#28085 )

2026-06-15 11:01:55 -07:00

distributed

[CAR] Let custom allreduce support VMM based allocation (#27593 )

2026-06-16 04:49:05 -07:00

dllm

Fix Req fill_len (fill_ids) having dual semantics by restricting to truncated/committed semantics (#26659 )

2026-06-08 14:52:48 +08:00

elastic_ep

chore: bump mooncake version to 0.3.11.post1 (#25989 )

2026-05-25 16:51:03 +08:00

entrypoints

chore: bump sglang-kernel version to 0.4.4 (#28572 )

2026-06-17 16:13:39 -07:00

eplb

LPLB: linear-programming load balancer for MoE expert parallelism (#24515 )

2026-06-16 10:19:42 -07:00

function_call

Support GLM-4.7 function calling via structural tags (#28149 )

2026-06-14 14:57:07 +08:00

grpc

[gRPC] Extract gRPC servicer into standalone package (#20478 )

2026-03-13 09:13:29 -07:00

hardware_backend

[MLX] Add Metal profiling hooks to server profiler (#28122 )

2026-06-17 13:06:20 -07:00

kv_canary

[Spec] Remove deprecated EAGLE v1 DRAFT_EXTEND forward mode (#28129 )

2026-06-13 13:00:21 -07:00

layers

[GDN][KDA][mem_cache] int8 checkpoint pool for the linear-attn prefix cache (#28185 )

2026-06-17 20:41:46 -07:00

lora

[LoRA] Fix chunked SGMV (csgmv) CUDA graph segment replay (#28371 )

2026-06-16 19:19:07 -07:00

managers

[GDN][KDA][mem_cache] int8 checkpoint pool for the linear-attn prefix cache (#28185 )

2026-06-17 20:41:46 -07:00

mem_cache

[GDN][KDA][mem_cache] int8 checkpoint pool for the linear-attn prefix cache (#28185 )

2026-06-17 20:41:46 -07:00

model_executor

Add get_parallel(): a structured accessor for parallel-topology state (#28567 )

2026-06-17 20:23:43 -07:00

model_loader

Add get_parallel(): a structured accessor for parallel-topology state (#28567 )

2026-06-17 20:23:43 -07:00

models

Add get_parallel(): a structured accessor for parallel-topology state (#28567 )

2026-06-17 20:23:43 -07:00

multimodal

[perf] Reuse a pooled HTTP session for multimodal URL downloads (#28056 )

2026-06-15 17:47:47 +08:00

multiplex

Move request-ingress methods to SchedulerRequestReceiver (#25610 )

2026-05-18 18:31:17 +08:00

observability

[observability] add Ray metric backend wrappers (#26252 )

2026-06-17 19:41:53 -07:00

parser

Fix Responses API request handling (#25881 )

2026-06-12 14:47:55 -07:00

platforms

Introduce CpuDeviceMixin and CpuSRTPlatform (#26385 )

2026-06-17 17:41:17 -07:00

plugins

Multi platform Plugin (#21388 )

2026-04-19 17:23:51 -07:00

ray

feat: add SGLANG_RAY_BUNDLE_INDICES for fine-grained Ray bundle index control (#24667 )

2026-05-30 02:19:50 -07:00

sampling

[Perf] Avoid per-decode-step host sync in min_new_tokens penalty (#28397 )

2026-06-15 23:32:28 -07:00

session

[perf] reduce overhead of fill_ids list reconstruction and decref (#27965 )

2026-06-14 00:41:11 -07:00

speculative

add dflash gemma4 support (#27471 )

2026-06-17 16:39:28 -07:00

state_capturer

Fix routed-experts device buffer overflow under DP attention (#26123 )

2026-05-30 19:11:06 -07:00

tokenizer

…

utils

fix: get_processor fails when --tokenizer-path lacks model config.json (#25643 )

2026-06-16 20:25:05 +03:00

weight_sync

…

constants.py

[AMD][MXFP4] Online MXFP4 quantization 1/N - dense and MOE models w. original BF16 weight (#18005 )

2026-06-03 12:55:24 -07:00

environ.py

Deepseek v4: support mixed dtype compression states (#27277 )

2026-06-17 00:56:41 -07:00

runtime_context.py

Add get_parallel(): a structured accessor for parallel-topology state (#28567 )

2026-06-17 20:23:43 -07:00

server_args_config_parser.py

fix(server): serialize nested dict config values as JSON (#28094 )

2026-06-12 19:27:31 -07:00

server_args.py

[GDN][KDA][mem_cache] int8 checkpoint pool for the linear-attn prefix cache (#28185 )

2026-06-17 20:41:46 -07:00