Han Lv
Software Developer at Lab of Mediate Energy Physics and Nuclear Technology
China
Summary
🤩
Rockstar🎓
Top SchoolHan Lv is a software developer based in China with 10 years of experience focused on back-end development and MLOps for machine learning model deployment and evaluation. At the Lab of Mediate Energy Physics and Nuclear Technology he builds and adapts deployment pipelines and inference integrations, bridging research models and production tooling. He has contributed to prominent open-source projects like OpenCompass—integrating TurboMind and lmdeploy pipelines for broad LLM evaluation—and MMDeploy, where he refined deployment configs and demos for detection and segmentation workloads. Han’s strengths include adapting SDKs, resolving compatibility issues across inference stacks, and streamlining demos and tests to improve reproducibility. Colleagues rely on him to translate cutting-edge model capabilities into stable, configurable deployment artifacts that work across diverse runtimes. An often-overlooked asset is his knack for simplifying complex config and pipeline mismatches, which speeds up adoption of new model backends.
10 years of coding experience
ustc
Github Skills (23)
apim10
pytorch10
deploying10
python10
evaluation10
convert10
converter10
tensorrt10
deeplearning-ai10
deep-learning10
api10
sdk10
computer-vision10
benchmark9
c-language9
Programming languages (5)
Github contributions (5)
open-mmlab/mmdeploy
Oct 2021 - Jan 2023
OpenMMLab Model Deployment Framework
Role in this project:
MLOps Engineer
MLOps EngineerContributions:11 releases, 1063 reviews, 84 commits in 1 year 3 months
Contributions summary:Han's commits primarily focus on integrating and adapting the MMDeploy framework for various machine learning tasks, including text detection, object detection, and image segmentation. They are responsible for modifying configurations to align with SDK model specifications, refactoring test code for improved structure, and updating demo examples to showcase the framework's functionality. Furthermore, the commits involve replacing and renaming specific dependencies like "tensorrt" for model deployment, indicating a focus on optimizing model deployment pipelines.
pytorchncnnpplnndeep-learningdeployment
open-compass/opencompass
Sep 2023 - Nov 2024
OpenCompass is an LLM evaluation platform, supporting a wide range of models (Llama3, Mistral, InternLM2,GPT-4,LLaMa2, Qwen,GLM, Claude, etc) over 100+ datasets.
Role in this project:Back-end Developer / MLOps Engineer
Back-end Developer / MLOps EngineerContributions:12 reviews, 7 PRs, 14 comments in 1 year 1 month
Contributions summary:Han primarily contributed to integrating TurboMind inference within the OpenCompass LLM evaluation platform. This involved utilizing both the TurboMind RPC API and Python API for inference, and included modifications to configuration files for model integration. They also adapted the codebase to be compatible with lmdeploy versions and integrated lmdeploy's pipeline API, enhancing the platform's inference capabilities. Furthermore, the user addressed several bugs and implemented updates, contributing to the overall functionality and stability of the project.
benchmarkchatgptevaluationlarge-language-modelllama2
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.