sglang

batch_invariant_ops

fix: add bias when enable mm fallback variant (#17690 )

2026-01-28 09:50:49 +08:00

batch_overlap

[MiMoV2Flash] [feat]: support two batch overlap (#17634 )

2026-02-02 14:03:57 -08:00

checkpoint_engine

[Ckpt Engine] feat: new sglang entrypoint support for update (#12216 )

2025-10-30 10:39:27 +08:00

compilation

Fix torch.__version__ for PEP440 (#15682 )

2026-01-29 11:55:13 -08:00

configs

model: support interns1-pro (#18145 )

2026-02-05 00:22:44 +08:00

connector

[Lint] Add python/sglang to ruff F401 checks and remove unused imports in files (#11685 )

2025-10-17 16:49:46 -07:00

constrained

Fix grammar sync across TP ranks (#17100 )

2026-01-15 18:38:01 -08:00

debug_utils

Support multiple engines for router simulation in schedule simulator (#16368 )

2026-01-04 11:46:05 +08:00

disaggregation

add hybrid model PD to NIXL connector (#16229 )

2026-02-06 15:05:36 -08:00

distributed

Make sure we always disable symm memory without dp padding (#18129 )

2026-02-04 19:58:28 +08:00

dllm

[DLLM] Implement initial dynamic batching for diffusion LLM (#14883 )

2026-01-17 16:48:15 +08:00

elastic_ep

[2/N] Added the core structure of elastic EP and the eplb algorithm with faulty rank (#10606 )

2025-10-22 01:13:31 -07:00

entrypoints

entrypoint: support passing spaces_between_special_tokens per request (#17939 )

2026-02-04 22:18:36 +08:00

eplb

Re-introduce the unit test of test_mooncake_ep_small (#16019 )

2026-01-12 17:01:24 +08:00

function_call

Gigachat 3 tool parser and tests (#14765 )

2026-02-02 22:28:34 -08:00

grpc

[gRPC] Add GetLoads RPC for comprehensive load metrics (#17087 )

2026-01-14 12:05:10 -08:00

hardware_backend/npu

Moving _alloc_extend_naive out of npu allocator (#18200 )

2026-02-04 02:09:55 -08:00

layers

[Qwen3Next] Optimize fused_sigmoid_gating_delta_rule_update_kernel (#18271 )

2026-02-07 11:59:42 +08:00

lora

throw error if got adapter with added_tokens (#18046 )

2026-02-05 23:55:43 +08:00

managers

fix kimi k2.5's moe gemm config init (#18064 )

2026-02-04 16:59:01 -08:00

mem_cache

optimize get_topk_ragged by fusing get k and k_scale triton kernel (#16043 )

2026-02-04 19:59:41 +08:00

metrics

[HiCache] feat: Add detailed cache hit breakdown for HiCache in sglext and Prometheus metrics (#17648 )

2026-02-03 11:45:35 -08:00

model_executor

model: support Step-3.5-Flash (#18084 )

2026-02-03 00:40:07 +08:00

model_loader

Fix cross-container HF download race condition in CI (#18328 )

2026-02-05 21:01:41 -08:00

models

Refactor(qwen3-vl) optimize position encoding interpolation (#16781 )

2026-02-05 10:26:35 -08:00

multimodal

model: support interns1-pro (#18145 )

2026-02-05 00:22:44 +08:00

multiplex

[Bugfix] fix the display error (inconsistent context) (#17699 )

2026-01-31 23:35:11 -08:00

parser

model: support Step-3.5-Flash (#18084 )

2026-02-03 00:40:07 +08:00

sampling

Revert "feat: reduce constrained-decoding overhead in TP" (#16845 )

2026-01-10 11:39:38 +08:00

speculative

model: support Step-3.5-Flash (#18084 )

2026-02-03 00:40:07 +08:00

tokenizer

[Piecewise Cuda Graph] rename, refactor and add more logging (#13675 )

2025-11-21 13:28:38 +08:00

tracing

SGLang Tracing: Improve root span attributes (#17008 )

2026-01-30 16:02:05 +08:00

utils

[XPU] Integrate MoE and minor improvements in XPU attention backend (#13561 )

2026-02-04 23:09:59 -08:00

weight_sync

[RL] Allow passing tensors of different dtypes for FlattenedTensorBucket (#13413 )

2025-11-19 17:24:48 +08:00

constants.py

Support releasing CUDA graph memory when paused (#7873 )

2025-10-28 14:40:50 +08:00

environ.py

[DeepGemm] Add a flag for fast warmup (#18111 )

2026-02-04 14:12:13 +08:00

server_args_config_parser.py

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

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

server_args.py

model: support Step-3.5-Flash (#18084 )

2026-02-03 00:40:07 +08:00