sglang

batch_invariant_ops

[deterministic inference] Move batch invariant pkg to sglang (#10695 )

2025-09-21 19:35:14 -07:00

configs

model: Support Hybrid Mamba2 NemotronHForCausalLM (nvidia/NVIDIA-Nemotron-Nano-9B-v2) (#10909 )

2025-10-09 00:37:38 +08:00

connector

Support loading weights from remote instance (#8215 )

2025-09-12 17:40:22 +08:00

constrained

[Auto Sync] Update base_grammar_backend.py, xgrammar_backen... (20250930) (#11115 )

2025-09-30 21:50:43 -07:00

debug_utils

Tiny improve dumper (#11132 )

2025-10-02 17:55:01 +08:00

disaggregation

fix(decode): adjust ServerArgs import to explicit module path (#11007 )

2025-10-08 01:27:50 -07:00

distributed

[Feat] Support Torch Symm Mem AllReduce (#10571 )

2025-10-05 13:55:19 -07:00

entrypoints

[router][grpc] disable health check generation and increase timeout (#11353 )

2025-10-08 19:23:08 -07:00

eplb

EPLB: prefer to use physical experts in the same node (#9849 )

2025-09-22 00:34:30 -07:00

function_call

Update tool parser and related documentation (#11223 )

2025-10-07 11:03:40 -07:00

grpc

[router] add get server info and get model info in grpc server (#11303 )

2025-10-07 08:36:52 -07:00

layers

[Generative Score API] Multi-Item scoring with custom attention mask. (#10979 )

2025-10-08 18:47:32 -07:00

lora

Fix LoRA support for multimodal models (VLMs) by implementing a consistent pattern for skipping vision components (#11261 )

2025-10-06 17:23:00 -07:00

managers

[Generative Score API] Multi-Item scoring with custom attention mask. (#10979 )

2025-10-08 18:47:32 -07:00

mem_cache

[DeepSeek-V3.2] Include indexer kv cache when estimating kv cache size (#11309 )

2025-10-08 23:59:46 -07:00

metrics

Minor fixes for server_args, parallel_state, and test_deterministic.py (#11159 )

2025-10-02 12:12:49 -07:00

model_executor

[DeepSeek-V3.2] Include indexer kv cache when estimating kv cache size (#11309 )

2025-10-08 23:59:46 -07:00

model_loader

Enable native ModelOpt quantization support (1/3) (#7149 )

2025-10-06 13:24:15 -07:00

models

model: Support Hybrid Mamba2 NemotronHForCausalLM (nvidia/NVIDIA-Nemotron-Nano-9B-v2) (#10909 )

2025-10-09 00:37:38 +08:00

multimodal

Support Dots.ocr model (#11071 )

2025-09-30 12:18:39 -07:00

parser

feat: add shortcut detection for multimodal templates in Jinja format (#11209 )

2025-10-06 04:13:17 +00:00

sampling

Remove sampling info events and overlap thread file (#11300 )

2025-10-07 21:34:25 +08:00

speculative

model: Support Hybrid Mamba2 NemotronHForCausalLM (nvidia/NVIDIA-Nemotron-Nano-9B-v2) (#10909 )

2025-10-09 00:37:38 +08:00

tokenizer

Add reasoning_effort param in TiktokenTokenizer.apply_chat_template (#9630 )

2025-08-26 00:44:20 -07:00

tracing

Fix metrics and request tracing (TimeStats) (#11123 )

2025-10-01 13:03:07 -07:00

utils

model: Support Hybrid Mamba2 NemotronHForCausalLM (nvidia/NVIDIA-Nemotron-Nano-9B-v2) (#10909 )

2025-10-09 00:37:38 +08:00

weight_sync

Tiny move files to utils folder (#11166 )

2025-10-03 22:40:06 +08:00

_custom_ops.py

[Feature] Integrate quick allreduce and select the best allreduce implementation (#6619 )

2025-07-24 20:48:42 -07:00

aio_rwlock.py

A better aio rwlock that guarantees the order (#2547 )

2024-12-22 15:44:32 -08:00

bench_utils.py

Revert "[1/2] Optimizations and refactors about quant kernel (#9534 )" (#10292 )

2025-09-10 18:24:23 -07:00

constants.py

Multi-Stage Awake: Support Resume and Pause KV Cache and Weights separately (#7099 )

2025-06-19 00:56:37 -07:00

custom_op.py

enable llama3.1-8B on xpu (#9434 )

2025-09-07 22:34:20 -07:00

environ.py

[CI] improve disaggregation CI. (#11264 )

2025-10-08 21:40:56 +08:00

host_shared_memory.py

Overlapped weight offload (#8034 )

2025-08-23 02:06:46 -07:00

offloader.py

Support offloading in fp8 (#9948 )

2025-09-14 01:14:28 -07:00

operations_strategy.py

[5/N] MoE Refactor: Update MoE parallelism arguments (#8658 )

2025-08-01 01:20:03 -07:00

operations.py

Add fp4 quantize before all-gather for Flashinfer cutlass MoE DP (max throughput) (#7667 )

2025-08-15 22:08:11 -07:00

server_args_config_parser.py

chore: Initial support for input config files (#10534 )

2025-09-24 14:45:52 -07:00

server_args.py

[DeepSeek-V3.2] Include indexer kv cache when estimating kv cache size (#11309 )

2025-10-08 23:59:46 -07:00

single_batch_overlap.py

Support single batch overlap (#10422 )

2025-10-02 18:04:36 +08:00

torch_memory_saver_adapter.py

[Feature] Option to save model weights to CPU when memory saver mode is enabled (#10873 )

2025-10-03 16:48:19 +08:00

two_batch_overlap.py

[Generative Score API] Multi-Item scoring with custom attention mask. (#10979 )

2025-10-08 18:47:32 -07:00

warmup.py

[Auto Sync] Update scheduler_profiler_mixin.py, rpd_utils.p... (20250916) (#10494 )

2025-09-16 17:02:20 -07:00