gingerXue e9d727cb92 [MUSA][7/N] Enhance CUDA / PyNccl wrapper to support MTLink connectivity detection (#17499) ...

Signed-off-by: jingzhi.xue <jingzhi.xue@mthreads.com>
Co-authored-by: jingzhi.xue <jingzhi.xue@mthreads.com>

2026-01-28 11:36:30 -08:00

..

batch_invariant_ops

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

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

batch_overlap

Overlap shared experts with deepep dispatch for single batch overlap on Blackwell (#17289)

2026-01-21 02:56:55 +08:00

checkpoint_engine

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

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

compilation

[Piecewise] Support PCG weak_ref_tensor cuda kernel on AMD (#17291)

2026-01-20 14:05:32 -08:00

configs

[GLM-OCR] Support GLM-OCR Model (#17582)

2026-01-26 22:24:00 -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

Introduce global alloc_len_per_decode & clean check decode memory (#15115)

2026-01-26 10:26:20 -08:00

distributed

[MUSA][7/N] Enhance CUDA / PyNccl wrapper to support MTLink connectivity detection (#17499)

2026-01-28 11:36:30 -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

[Bug Fix] Fix reasoning parser when continue_final_message=true (#17065)

2026-01-27 14:04:44 +08:00

eplb

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

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

function_call

model: step3-vl-10b (#17513)

2026-01-22 23:15:08 +08:00

grpc

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

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

hardware_backend/npu

[NPU] NZ for non-quantized MOE, Qwen3 MOE double memory consumption fix (#15904)

2026-01-29 00:55:08 +08:00

layers

[NPU] NZ for non-quantized MOE, Qwen3 MOE double memory consumption fix (#15904)

2026-01-29 00:55:08 +08:00

lora

add documentation example for LoRA overlap loading and cleanup unused function (#17464)

2026-01-24 15:33:16 +08:00

managers

use shared memory for multimodal feature transport between Tokenizer and Scheduler (#16402)

2026-01-27 11:01:08 -08:00

mem_cache

[HiCache][HA 1/N] Support HiCache storage runtime attach/detach (#15892)

2026-01-26 19:33:19 -08:00

metrics

Code clean up for fp8 quantization (#16982)

2026-01-13 12:38:39 -08:00

model_executor

[chore]: improve time tracing of model loading process (#15426)

2026-01-26 19:04:25 -08:00

model_loader

[chore]: improve time tracing of model loading process (#15426)

2026-01-26 19:04:25 -08:00

models

[NPU] NZ for non-quantized MOE, Qwen3 MOE double memory consumption fix (#15904)

2026-01-29 00:55:08 +08:00

multimodal

[GLM-OCR] Support GLM-OCR Model (#17582)

2026-01-26 22:24:00 -08:00

multiplex

Fix deprecation warning of sre_parse for python 3.13 (#16247)

2025-12-31 21:25:09 -08:00

parser

[Bug Fix] Fix reasoning parser when continue_final_message=true (#17065)

2026-01-27 14:04:44 +08:00

sampling

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

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

speculative

Introduce global alloc_len_per_decode & clean check decode memory (#15115)

2026-01-26 10:26:20 -08:00

tokenizer

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

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

tracing

SGLang Tracing: fix attribute errors (header extraction & bootstrap span closing) (#15693)

2025-12-27 22:44:31 -06:00

utils

[GLM-OCR] Support GLM-OCR Model (#17582)

2026-01-26 22:24:00 -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

[NIXL] Add custom NIXL backend selection for KVManager (#17146)

2026-01-26 14:35:38 +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] Add support for EXAONE-4.0 Model (#8205)

2026-01-27 14:08:24 +08:00