Tamir Baydasov aeca7d348c [3/N] Quantization Refactor: ModelSlim MoE schemes (#17993) ...

Co-authored-by: ronnie_zheng <zl19940307@163.com>

2026-02-17 21:38:27 +03:00

..

batch_invariant_ops

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

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

batch_overlap

[TBO] fix cuda graph intermittently becomes disabled bug (#18320)

2026-02-16 22:18:57 -08:00

checkpoint_engine

update pre-commit config (#18860)

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

compilation

update pre-commit config (#18860)

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

configs

[Model] Add Qwen3ForRewardModel and fix Qwen3ForSequenceClassification (#17992)

2026-02-16 19:44:41 +08:00

connector

…

constrained

…

debug_utils

Support dumping gradients, parameters, lazy values (#18881)

2026-02-16 13:34:06 +08:00

disaggregation

refactor FAKE transfer backend and remove --disaggregation-decode-enable-fake-auto parameter (#18345)

2026-02-16 17:27:02 +03:00

distributed

update pre-commit config (#18860)

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

dllm

Fix prefill stats for dllm (#18632)

2026-02-12 01:00:30 +08:00

elastic_ep

…

entrypoints

update pre-commit config (#18860)

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

eplb

…

function_call

update pre-commit config (#18860)

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

grpc

[gRPC] Fix scheduler startup broken by context parallel refactor (#18933)

2026-02-17 08:52:11 -08:00

hardware_backend/npu

update pre-commit config (#18860)

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

layers

[3/N] Quantization Refactor: ModelSlim MoE schemes (#17993)

2026-02-17 21:38:27 +03:00

lora

throw error if got adapter with added_tokens (#18046)

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

managers

[misc] adding metadata field in UpdateWeightFromDiskReqInput (#18821)

2026-02-17 12:14:15 +08:00

mem_cache

Nsa trtllm mla sparse fp8 support with Deepseek v3.2 NVFP4 (#18389)

2026-02-16 09:29:54 +08:00

metrics

update pre-commit config (#18860)

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

model_executor

Nsa trtllm mla sparse fp8 support with Deepseek v3.2 NVFP4 (#18389)

2026-02-16 09:29:54 +08:00

model_loader

update pre-commit config (#18860)

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

models

[PCG] support piecewise cuda graph for kimi-linear model (#18849)

2026-02-17 23:31:12 +08:00

multimodal

Fix GLM-4V processor registration when glm_ocr is unavailable (#18885)

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

multiplex

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

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

parser

update pre-commit config (#18860)

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

sampling

…

speculative

refactor context parallel state (#17213)

2026-02-13 23:18:17 +08:00

tokenizer

…

tracing

SGLang Tracing: Improve root span attributes (#17008)

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

utils

Fix libnuma.so does not exsit (#15355)

2026-02-16 00:37:50 +08:00

weight_sync

…

constants.py

…

environ.py

Flip dumper to disable by default and refactor environment handling (#18878)

2026-02-16 13:29:32 +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

refactor FAKE transfer backend and remove --disaggregation-decode-enable-fake-auto parameter (#18345)

2026-02-16 17:27:02 +03:00