b8zhong 74a166cb86 [Fix] Only add SM90 and SM100 to check for auto-enabling TRT Allreduce Fusion (#16283)

2026-01-03 11:43:17 +08:00

..

batch_invariant_ops

…

batch_overlap

[Feature] Xiaomi MiMo-V2-Flash day0 support (#15207)

2025-12-19 11:40:07 +08:00

checkpoint_engine

…

compilation

Custom All Reduce for Piecewise Cuda Graph (#15356)

2025-12-25 23:55:15 +08:00

configs

[NemotronH] Add latent MoE support (#16227)

2026-01-02 22:08:58 +08:00

connector

…

constrained

Adding tool calling and reasoning parser support for Intern-S1 (#14866)

2025-12-16 00:01:58 -08:00

debug_utils

Tiny fix single-gpu dumper and add tests for dumper (#16285)

2026-01-02 13:30:09 +08:00

disaggregation

Move log_prefill_stats_late to correct location in PP mode (#15946)

2025-12-31 12:00:24 +08:00

distributed

[distributed] Clean up MoE groups in destroy_model_parallel (#15345)

2026-01-02 14:45:14 +08:00

dllm

Tiny extract ModelRunnerOutput (#15400)

2025-12-18 22:18:45 +08:00

elastic_ep

…

entrypoints

[Auto Sync] Update engine.py (20260102) (#16317)

2026-01-02 17:29:40 -08:00

eplb

Support EPLB balancedness prometheus metric without GPU->CPU synchronize (#15401)

2025-12-18 22:24:23 +08:00

function_call

Fix DeepSeekV31's structural tag trigger (#13394)

2025-12-31 21:13:52 +08:00

grpc

[model-gateway] Add classification model support infrastructure (#16061)

2025-12-29 08:34:05 -08:00

hardware_backend/npu

[DeepSeek v3.2] opt Context Parallelism: support fused moe, multi batch and fp8 kvcache (#13959)

2026-01-02 23:49:14 +08:00

layers

[DeepSeek v3.2] opt Context Parallelism: support fused moe, multi batch and fp8 kvcache (#13959)

2026-01-02 23:49:14 +08:00

lora

[LoRA] Torch native backend: rework implementation and updated tests (#15187)

2025-12-30 11:48:51 +08:00

managers

[DeepSeek v3.2] opt Context Parallelism: support fused moe, multi batch and fp8 kvcache (#13959)

2026-01-02 23:49:14 +08:00

mem_cache

Support swa allocator page size > 1 (#16296)

2026-01-03 08:49:47 +08:00

metrics

Add LoRA metrics for potential auto scaling (#15149)

2025-12-24 12:13:52 -08:00

model_executor

[AMD] feat: add DLLM support for AMD GPUs with LLaDA2 testing (#15560)

2026-01-03 10:41:11 +08:00

model_loader

Download missing shards in model weights files when not in CI (#16211)

2025-12-31 20:42:34 +08:00

models

[Tiny]Remove progress bar for fp8 ue8m0 quant when unneeded (#16177)

2026-01-03 10:54:12 +08:00

multimodal

[VLM] feat: true on policy for vlm + fsdp (#14636)

2026-01-01 16:54:39 -08:00

multiplex

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

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

parser

fix: prevent points regex from matching checkpoints/endpoints (#15120)

2025-12-17 12:38:18 -08:00

sampling

Support returning customized info (#16144)

2026-01-02 15:32:18 +08:00

speculative

Eagle: GPT-OSS Eagle v2 support (#14920)

2025-12-30 14:23:07 -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

[Refactor] Clean up custom op (#15995)

2026-01-01 21:41:56 +08:00

weight_sync

…

constants.py

…

environ.py

Add host tensor allocator for memory_pool_host and support Mooncake standalone storage (#14873)

2025-12-28 16:28:50 -08:00

server_args_config_parser.py

Fix parse args from file(#13911) (#14085)

2026-01-01 11:37:33 +08:00

server_args.py

[Fix] Only add SM90 and SM100 to check for auto-enabling TRT Allreduce Fusion (#16283)

2026-01-03 11:43:17 +08:00