sglang

arg_groups

introduce arg_groups/ with nemotron_h hook (#24328 )

2026-05-03 16:28:11 -07:00

batch_invariant_ops

Register aten::rms_norm and aten::mm.dtype in batch invariant mode (#24459 )

2026-05-05 17:21:34 -07:00

batch_overlap

[bugfix] Support MIXED forward mode in TBO splitter for DP attention (#24241 )

2026-05-01 16:01:23 -07:00

checkpoint_engine

update pre-commit config (#18860 )

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

compilation

Multi platform Plugin (#21388 )

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

configs

add indexer-topk capture (V3.2 NSA + infra) (#24392 )

2026-05-05 15:05:15 -07:00

connector

Remove sync points in mamba cache + prefill cudagraph plumbing for DP (#19639 )

2026-03-02 15:03:42 +08:00

constrained

[Fix] eagle/eagle3 speculative decoding conflicts with xgrammar in NPU (#20989 )

2026-04-15 23:34:23 -07:00

debug_utils

Add e2e test with log snapshot in dumper grafter (#24513 )

2026-05-06 17:00:13 +08:00

disaggregation

[PD] Prevent update_status to Failed from cleared entries (#24539 )

2026-05-06 23:32:04 +08:00

distributed

Refactor: decouple segment tracking from comm registration (#21392 )

2026-05-06 17:07:58 +08:00

dllm

Decouple scheduler log printing from metrics collection (#20107 )

2026-03-07 22:09:10 -08:00

elastic_ep

[6/N] (Elastic EP) Recover failed ranks (#15771 )

2026-04-28 00:44:26 -07:00

entrypoints

[Kernel] Deprecate DeepGemm in sgl kernel and apply custom wheel sgl-deep-gemm (#24268 )

2026-05-06 18:59:01 -07:00

eplb

Replace hardcoded CUDA device with get_device() for XPU support (#13599 )

2026-05-01 07:13:46 +08:00

function_call

feat: use structural tags to enable strict tool calling and reasoning for more models (#21722 )

2026-05-04 02:30:28 -07:00

grpc

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

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

hardware_backend

move topk capturers to srt/state_capturer/ (#24450 )

2026-05-05 15:54:01 -07:00

layers

[Kernel] Deprecate DeepGemm in sgl kernel and apply custom wheel sgl-deep-gemm (#24268 )

2026-05-06 18:59:01 -07:00

lora

[LoRA] Use deterministic lora_id for --lora-paths so multi-node ranks agree (#24555 )

2026-05-06 22:20:15 -07:00

managers

Improve metrics, observability, and PD deploy tooling (#24521 )

2026-05-06 11:27:35 -07:00

mem_cache

Support swa HiCache for unified radix cache (#23391 )

2026-05-06 22:19:25 +08:00

model_executor

Support getting checksums in weight checker (#24537 )

2026-05-06 22:59:28 +08:00

model_loader

nextn subclass owns post_load_weights is_nextn (#24333 )

2026-05-03 22:04:44 -07:00

models

move topk capturers to srt/state_capturer/ (#24450 )

2026-05-05 15:54:01 -07:00

multimodal

[Feature] Xiaomi MiMo-V2.5 day0 support (#23811 )

2026-05-01 00:02:26 +08:00

multiplex

[mem] Flatten memory checkers into composable per-pool invariant checks (#22562 )

2026-04-11 02:56:22 -07:00

observability

Improve metrics, observability, and PD deploy tooling (#24521 )

2026-05-06 11:27:35 -07:00

parser

[Feature] Xiaomi MiMo-V2.5 day0 support (#23811 )

2026-05-01 00:02:26 +08:00

platforms

Clean up noisy startup warnings from third-party deps (#23669 )

2026-04-27 03:10:46 -07:00

plugins

Multi platform Plugin (#21388 )

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

ray

[Ray] Bind scheduler actors to GPU-local NUMA node (#22989 )

2026-04-16 14:52:15 -07:00

sampling

Fix /generate endpoint crash when sampling params contain null values (#23401 )

2026-04-22 09:56:10 -07:00

session

[session] fix mamba pool leak in StreamingSession.release_session + plumb idle leak check (#23496 )

2026-05-02 11:38:08 +08:00

speculative

fix(req_pool): bump pool.size to match actual tensor row count after #24243 (#24439 )

2026-05-05 16:58:26 -07:00

state_capturer

[R3] Avoid implicit CUDA sync in routed experts DP slicing (#24550 )

2026-05-06 18:37:36 -07:00

tokenizer

…

utils

[Bug Fix] Fix RunAI streamer: corrupted weights, missing quant init, and broken URIs for multimodal models (#22715 )

2026-05-06 19:20:04 -07:00

weight_sync

…

constants.py

Clean up detokenizer and remove dead multimodal_gen code (#21588 )

2026-03-28 21:44:40 -07:00

environ.py

Rename SGLANG_USE_JIT_ALL_REDUCE to SGLANG_OPT_USE_CUSTOM_ALL_REDUCE_V2 (#24297 )

2026-05-02 23:59:46 -07:00

server_args_config_parser.py

[ConfigArgumentMerger] Improve ConfigArgumentMerger compatibility with external callers (#17051 )

2026-01-16 15:32:42 +08:00

server_args.py

[LoRA] Use deterministic lora_id for --lora-paths so multi-node ranks agree (#24555 )

2026-05-06 22:20:15 -07:00