京东大模型算法工程师
社招全职算法开发岗地点:北京状态:招聘
任职要求
1. 教育背景: 计算机科学、人工智能或相关 STEM 领域硕士及以上学历,具备扎实的理论基础; 2. 技术能力: * 优秀的基础: 拥有深厚的计算机科学理论根基与扎实的算法功底; * 编程实力: 具备卓越的编程能力和良好的工程实践习惯; * 引擎熟悉度: 熟练掌握 vLLM, SGLang, TensorRT-LLM 等主流大模型推理引擎之一,理解其核心原理; 3、经验优先: * 在自然语言处理、多模态大模型、扩散模型、图神经网络(GNN)等领域有研究、技术开发或实际落地经验; * 作为核心骨干在国际顶会(NeurIPS, ICML, ICLR,CVPR, ACL等)发表过高质量论文; * 在知名开源项目(如Hugging Face Transformers, DeepSpeed,vLLM,SGLang,TensorRT-LLM等)中有显著贡献; * 在权威人工智能相关竞赛(如Kaggle、天池)中获得优异名次; 4、软性素质: * 沟通协作: 具备出色的沟通表达能力和高效的团队协作精神; * 技术追求: 对技术有强烈的好奇心与钻研精神,追求卓越; * 解决问题: 具备优秀的自驱力,能主动发现问题、分析问题并创造性地解决问题。 符合京东价值观:客户为先、创新、拼搏、担当、感恩、诚信。
工作职责
1. 研发高性能推理算法: 设计并实现自注意力机制优化、并行推理、负载均衡、弹性容量等无损推理服务优化算法,提升服务效率与稳定性; 2. 探索轻量化推理技术: 深入研究和应用有损推理加速算法,包括但不限于知识蒸馏、模型量化、网络剪枝、KV-Cache压缩等,实现模型的高效部署; 3. 聚焦软硬件一体优化策略: 从计算图优化、算子融合、计算通信重叠、专家并行、vGPU虚拟化等多个维度入手,显著提升端到端推理性能。
包括英文材料
学历+
算法+
https://roadmap.sh/datastructures-and-algorithms
Step by step guide to learn Data Structures and Algorithms in 2025
https://www.hellointerview.com/learn/code
A visual guide to the most important patterns and approaches for the coding interview.
https://www.w3schools.com/dsa/
大模型+
https://www.youtube.com/watch?v=xZDB1naRUlk
You will build projects with LLMs that will enable you to create dynamic interfaces, interact with vast amounts of text data, and even empower LLMs with the capability to browse the internet for research papers.
https://www.youtube.com/watch?v=zjkBMFhNj_g
vLLM+
https://www.newline.co/@zaoyang/ultimate-guide-to-vllm--aad8b65d
vLLM is a framework designed to make large language models faster, more efficient, and better suited for production environments.
https://www.youtube.com/watch?v=Ju2FrqIrdx0
vLLM is a cutting-edge serving engine designed for large language models (LLMs), offering unparalleled performance and efficiency for AI-driven applications.
SGLang+
[英文] Install SGLang
https://docs.sglang.ai/get_started/install.html
SGLang is a fast serving framework for large language models and vision language models.
https://github.com/sgl-project/sgl-learning-materials
TensorRT+
https://docs.nvidia.com/deeplearning/tensorrt/latest/getting-started/quick-start-guide.html
This TensorRT Quick Start Guide is a starting point for developers who want to try out the TensorRT SDK; specifically, it demonstrates how to quickly construct an application to run inference on a TensorRT engine.
推理引擎+
https://www.youtube.com/watch?v=_dvk75LEJ34
https://www.youtube.com/watch?v=XtT5i0ZeHHE
NLP+
https://www.youtube.com/watch?v=fNxaJsNG3-s&list=PLQY2H8rRoyvzDbLUZkbudP-MFQZwNmU4S
Welcome to Zero to Hero for Natural Language Processing using TensorFlow!
https://www.youtube.com/watch?v=R-AG4-qZs1A&list=PLeo1K3hjS3uuvuAXhYjV2lMEShq2UYSwX
Natural Language Processing tutorial for beginners series in Python.
https://www.youtube.com/watch?v=rmVRLeJRkl4&list=PLoROMvodv4rMFqRtEuo6SGjY4XbRIVRd4
The foundations of the effective modern methods for deep learning applied to NLP.
GNN+
https://distill.pub/2021/gnn-intro/
Neural networks have been adapted to leverage the structure and properties of graphs.
https://gnn.seas.upenn.edu/
Graph Neural Networks (GNNs) are information processing architectures for signals supported on graphs.
https://www.ibm.com/think/topics/graph-neural-network
Graph neural networks (GNNs) are a deep neural network architecture that is popular both in practical applications and cutting-edge machine learning research.
NeurIPS+
https://neurips.cc/
ICML+
https://icml.cc/
ICLR+
https://iclr.cc/
CVPR+
https://cvpr.thecvf.com/
DeepSpeed+
https://www.youtube.com/watch?v=pDGI668pNg0
Kaggle+
[英文] Kaggle Learn
https://www.kaggle.com/learn
Gain the skills you need to do independent data science projects.
相关职位
社招1年以上算法开发岗
1、参与生成式大模型能力构建;不局限于模型设计、prompt优化、预训练、模型推理加速、其他能力建设等; 2、采用最先进的并行处理和分布式学习技术,制定并执行性能优化策略,显著提升大型语言模型的训练速度和推理能力,例如跟进DeepSeek R1技术架构等,确保技术行业领先; 3、推进大模型技术在京东物流各个业务场景落地,包括不限于智能问答、智能数据分析、智能决策以及Computer Use等,助力业务流程优化,增质提效; 4、深度探索大语言模型方向,保持技术领先优势,推动京东物流在行业内树立高效、精准的大模型/多模态大模型应用标杆,并取得业务收益。
更新于 2025-06-09
社招大模型
1、探索新一代大语言模型基座架构,完成扩散模型(diffusion model)在大语言模型的重塑,突破逐个token预测的方式,实现高效的推理模式,探索全新scaling law; 2、实现大模型训练的数据清洗、合成和评估;设计和实现大模型训练的AI Infra框架。
更新于 2025-09-05