Skip to content
View CUHKSZzxy's full-sized avatar
:octocat:
Focusing
:octocat:
Focusing
36 followers · 30 following

Achievements

Achievement: Pair Extraordinairex2Achievement: QuickdrawAchievement: Pull Sharkx2Achievement: Arctic Code Vault Contributor

Achievements

Achievement: Pair Extraordinairex2Achievement: QuickdrawAchievement: Pull Sharkx2Achievement: Arctic Code Vault Contributor

Block or report CUHKSZzxy

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
CUHKSZzxy/README.md

Hi there 👋

Stats

Open-source Work

Focused on LLM/VLM serving systems.

Selected contributions:

#PR Status Summary
LMDeploy model support 🟣 #4575 Intern-S2-Preview; #4411 Qwen3-Omni; #4318 Intern-S1-Pro
#4093 Qwen3-VL; #3863 GLM-4.5; #3846 GLM-4.1V
#3315 Qwen3/MoE; #3194 Qwen2.5-VL; #3149 DeepSeek-VL2.
LMDeploy #4644 🟢 Unify interleaved MRoPE handling. [agent-assisted]
LMDeploy #4640 🟢 Add multimodal preprocessing metrics. [agent-assisted]
LMDeploy #4582 🟢 Add OpenAI Responses-compatible API endpoint. [agent-assisted]
LMDeploy #4563 🟣 Enable FP8 KV-cache quantization. [agent-assisted]
LMDeploy #4531 🟣 Handle mixed-modality serving paths. [agent-assisted]
LMDeploy #4452 🟣 Update draft model parameters for RL.
LMDeploy #4360 🟣 Handle video inputs.
LMDeploy #3534 🟣 Add serving metrics.
vLLM #42705 🟣 Support Intern-S2-Preview (co-authored).
vLLM #33636 🟣 Support Intern-S1-Pro.
SGLang #9299 🟣 Support InternS1-Mini.
SGLang #8350 🟣 Support Intern-S1 (co-authored).

Selected bug fixes:

#PR Status Summary
LMDeploy #4603 🟣 Fix VLM feature memory overhead. [agent-assisted]
LMDeploy #4583 🟣 Fix split multimodal tensor compaction.
LMDeploy #4084 🟣 Fix expert-parallel deployment issues.
LMDeploy #4003 🟣 Fix InternVL Flash long-context accuracy.

Status: 🟣 merged, 🟢 open.

[agent-assisted] marks work where I actively explored implementation possibilities with coding agents.

Pinned Loading

  1. InternLM/lmdeploy InternLM/lmdeploy Public

    LMDeploy is a toolkit for compressing, deploying, and serving LLMs.

    Python 7.9k 700

  2. vllm-project/vllm vllm-project/vllm Public

    A high-throughput and memory-efficient inference and serving engine for LLMs

    Python 82.2k 17.8k

  3. InternLM/Intern-S1 InternLM/Intern-S1 Public

    A Scientific Multimodal Foundation Model

    809 46

  4. vllm-project/recipes vllm-project/recipes Public

    Common recipes to run vLLM

    JavaScript 833 292