Baizhou Zhang 39008955ff Revert "[AMD][MORI] Fix MTP crash with FP4/FP8 dispatch and add NEXTN dispatch env vars." (#20602)

2026-03-14 12:12:42 -07:00

..

batch_invariant_ops

…

batch_overlap

[Feature] Integrate Elastic NIXL-EP into SGLang (#19248)

2026-03-11 17:37:43 +08:00

checkpoint_engine

…

compilation

Piecewise Cuda Graph set default (#16331)

2026-03-02 23:18:07 +08:00

configs

[Model] Support Nemotron 3 Super NVFP4 (#20407)

2026-03-14 00:56:26 -07:00

connector

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

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

constrained

[Minor] fix type annotations and invalid method calls in constrained … (#20132)

2026-03-12 16:42:46 -07:00

debug_utils

Enhance error resilience in dump comparator (#19685)

2026-03-02 19:08:35 +08:00

disaggregation

[NPU] Support mamba cache transfer for NPU (#20364)

2026-03-12 12:49:21 +08:00

distributed

[Feature] Integrate Elastic NIXL-EP into SGLang (#19248)

2026-03-11 17:37:43 +08:00

dllm

Decouple scheduler log printing from metrics collection (#20107)

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

elastic_ep

[4/N] (Elastic EP) Back up Expert Weights in DRAM (#17374)

2026-02-27 15:59:13 +08:00

entrypoints

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

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

eplb

[Feature] Integrate Elastic NIXL-EP into SGLang (#19248)

2026-03-11 17:37:43 +08:00

function_call

[Tool Call] Stream DeepSeek-V3.2 function call parameters in JSON format. (#16091)

2026-03-03 01:46:29 -08:00

grpc

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

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

hardware_backend/npu

[NPU] [Quantization] w4a4 MoE layer support (#18924)

2026-03-11 16:52:35 +03:00

layers

Revert "[AMD][MORI] Fix MTP crash with FP4/FP8 dispatch and add NEXTN dispatch env vars." (#20602)

2026-03-14 12:12:42 -07:00

lora

[NPU] make torch_native lora backend a little bit faster (#17228)

2026-03-07 20:14:46 +03:00

managers

[PP] Fix recv tensor dict potential race condition (#20341)

2026-03-14 13:35:01 +08:00

mem_cache

[bugfix] Add prev_prefix_len parameter to HiMambaRadixCache's _insert_helper() (#20539)

2026-03-14 09:54:14 +08:00

model_executor

Enable Piecewise CUDA Graph for NemotronH Hybrid (Mamba+Attention) Models (#19903)

2026-03-12 09:16:38 +08:00

model_loader

[Model] Support Nemotron 3 Super NVFP4 (#20407)

2026-03-14 00:56:26 -07:00

models

Revert "[AMD][MORI] Fix MTP crash with FP4/FP8 dispatch and add NEXTN dispatch env vars." (#20602)

2026-03-14 12:12:42 -07:00

multimodal

[NPU][QwenVL] Support qwen image preprocess on npu (#20189)

2026-03-11 15:03:08 +08:00

multiplex

Revert "[Scheduler] Decouple maybe_send_health_check_signal from process_batch_result" (#20259)

2026-03-10 01:58:48 -07:00

observability

Fix hicache log metrics (#20504)

2026-03-13 16:29:58 +08:00

parser

rename nemotron reasoning parser (#19865)

2026-03-05 11:27:07 -08:00

ray

Add Ray actor support for scheduler process management (DP=1) (#17684)

2026-03-05 13:21:23 -08:00

sampling

…

speculative

[AMD] Support speculative decoding v2 for aiter backend on ROCm/HIP (#17450)

2026-03-11 17:01:01 -07:00

tokenizer

…

utils

[Utils] Add NetworkAddress abstraction for IPv6-safe address handling (#20306)

2026-03-11 00:07:37 -07:00

weight_sync

…

constants.py

…

environ.py

[quant] Ignore FP8 quantization layers (#20340)

2026-03-13 13:59:39 +08:00

server_args_config_parser.py

…

server_args.py

[Model] Support Nemotron 3 Super NVFP4 (#20407)

2026-03-14 00:56:26 -07:00