haowen-han b992828ad2 fix: fix bug on kimi2.5 with dp2 and tp4 (#18604) ...

Co-authored-by: hanhaowen <hanhaowen@baidu.com>

2026-02-15 16:32:13 +08:00

..

batch_invariant_ops

…

batch_overlap

…

checkpoint_engine

…

compilation

[PCG] GPT OSS Triton Kernel Support (#18405)

2026-02-12 09:23:55 +08:00

configs

Fix/qwen3 5 amd rope cutedsl fallback (#18753)

2026-02-14 22:09:44 -08:00

connector

…

constrained

…

debug_utils

…

disaggregation

[PD-Disagg] Fix double free when prebuilt batch is aborted. (#18822)

2026-02-13 17:46:35 -08:00

distributed

refactor context parallel state (#17213)

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

dllm

Fix prefill stats for dllm (#18632)

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

elastic_ep

…

entrypoints

Fix dsv32 encode_messages (#18126)

2026-02-14 16:44:13 +08:00

eplb

…

function_call

…

grpc

[Auto Sync] Update grpc_request_manager.py, tokenizer_manag... (20260214) (#18838)

2026-02-14 18:12:32 -08:00

hardware_backend/npu

[NPU] support model skywork-reward-gemma2-2-27B-v0.2 (#16947)

2026-02-11 15:34:53 +08:00

layers

fix: fix bug on kimi2.5 with dp2 and tp4 (#18604)

2026-02-15 16:32:13 +08:00

lora

throw error if got adapter with added_tokens (#18046)

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

managers

[Auto Sync] Update grpc_request_manager.py, tokenizer_manag... (20260214) (#18838)

2026-02-14 18:12:32 -08:00

mem_cache

[Env] centralize hicache vars in environ.py (#17204)

2026-02-13 18:02:31 -08:00

metrics

Add cache_config_info metric. (#17273)

2026-02-09 16:09:09 -08:00

model_executor

Model: Support IBM Granite (Dense/Mamba + MoE) (#18040)

2026-02-15 11:24:41 +08:00

model_loader

[Auto Sync] Update loader.py, weight_utils.py (20260213) (#18779)

2026-02-13 12:22:50 -08:00

models

Model: Support IBM Granite (Dense/Mamba + MoE) (#18040)

2026-02-15 11:24:41 +08:00

multimodal

model: support Qwen3.5 (#18489)

2026-02-10 00:27:59 +08:00

multiplex

…

parser

Fix dsv32 encode_messages (#18126)

2026-02-14 16:44:13 +08:00

sampling

…

speculative

refactor context parallel state (#17213)

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

tokenizer

…

tracing

…

utils

Model: Support IBM Granite (Dense/Mamba + MoE) (#18040)

2026-02-15 11:24:41 +08:00

weight_sync

…

constants.py

…

environ.py

Kernel: optimize decoding metadata in NSA multi-spec backend with fused kernels (#17554)

2026-02-14 16:40:15 +08:00

server_args_config_parser.py

…

server_args.py

Model: Support IBM Granite (Dense/Mamba + MoE) (#18040)

2026-02-15 11:24:41 +08:00