Xiaoyu Zhang 03b835e7d1 Refactor tuning block wise kernel and opt Qwen/Qwen3-VL-32B-Instruct-FP8 (#14141)

2025-12-08 09:24:58 +08:00

..

batch_invariant_ops

Support fast gemm when in batch invariant DeepGEMM fallback (#13259)

2025-11-15 16:34:15 +08:00

batch_overlap

[Misc]Register and refactor some environs for dpsk-fp4 and DeepEp (#14538)

2025-12-06 12:29:16 -08:00

checkpoint_engine

…

compilation

[Auto Sync] Update backend.py (20251130) (#14153)

2025-11-30 22:15:02 -08:00

configs

Add Mistral Large 3 Eagle Support (#14466)

2025-12-05 23:11:41 +08:00

connector

…

constrained

fix "GrammarMatcher has terminated after accepting the stop token, but is trying to find the next token mask" when both reasoning and spec are enabled (#14464)

2025-12-06 06:15:22 -08:00

debug_utils

Super tiny add more info in dumper (#14145)

2025-11-30 09:07:39 +08:00

disaggregation

Enhance prefill PP node robustness (#14494)

2025-12-06 18:00:54 -08:00

distributed

Move custom_ops under layers; move _custom_ops.py → custom_all_reduce_ops.py (#14326)

2025-12-03 10:33:37 -08:00

dllm

[DLLM] feat: Add threshold based parallel decoding support (#14412)

2025-12-07 18:25:33 +08:00

elastic_ep

…

entrypoints

chore: bump sgl-kernel version to 0.3.18.post3 (#14518)

2025-12-06 13:15:16 -08:00

eplb

[NPU][1/N] NPU basic functions refactor and new modelslim quant type (#13359)

2025-12-04 16:15:31 +08:00

function_call

feat: DeepSeek new v3.2 encoding (#14249)

2025-12-02 11:41:05 -08:00

grpc

[misc] add model arch and type to server info and use it for harmony (#14456)

2025-12-05 06:51:00 -08:00

hardware_backend/npu

[BugFix] fix prefixcache performance and accuracy on ascend (#13573)

2025-12-08 02:16:20 +08:00

layers

Refactor tuning block wise kernel and opt Qwen/Qwen3-VL-32B-Instruct-FP8 (#14141)

2025-12-08 09:24:58 +08:00

lora

[NPU]LoRA: Adding Torch Native backend (#14132)

2025-12-08 02:16:07 +08:00

managers

[BugFix] fix prefixcache performance and accuracy on ascend (#13573)

2025-12-08 02:16:20 +08:00

mem_cache

[BugFix] fix prefixcache performance and accuracy on ascend (#13573)

2025-12-08 02:16:20 +08:00

metrics

[Auto Sync] Add max_total_num_tokens metric: Update scheduler_metrics_mixin.py, collector.py (20251202) (#14256)

2025-12-01 16:34:34 -08:00

model_executor

Add small model test for spec v2 + dp + trtllm_mla (#14576)

2025-12-07 23:55:00 +08:00

model_loader

Fix safetensors validation to catch corruption after download (#14465)

2025-12-05 16:04:00 -08:00

models

Tiny remove wrong import from python.sglang (#14577)

2025-12-07 22:07:49 +08:00

multimodal

ministral3 (#14251)

2025-12-04 14:31:26 -08:00

multiplex

…

parser

…

sampling

[Minor] update docs (#14212)

2025-12-01 02:33:58 -08:00

speculative

support mtp with deepseek r1 nvfp4 model (#13115)

2025-12-06 00:45:54 -08:00

tokenizer

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

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

tracing

[minor] remove debug code in python/sglang/srt/compilation/weak_ref_tensor_jit.py (#13235)

2025-11-14 12:22:24 -08:00

utils

support mtp with deepseek r1 nvfp4 model (#13115)

2025-12-06 00:45:54 -08:00

weight_sync

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

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

constants.py

…

custom_op.py

…

environ.py

[Misc]Register and refactor some environs for dpsk-fp4 and DeepEp (#14538)

2025-12-06 12:29:16 -08:00

server_args_config_parser.py

…

server_args.py

[NPU]LoRA: Adding Torch Native backend (#14132)

2025-12-08 02:16:07 +08:00