Foundational Model Research Engineer

Responsibilities:

  • Provide critical infrastructure to scale cutting-edge research into industry-leading next-generation models, including large-scale data acquisition, reinforcement learning environment development, and optimization of training efficiency.
  • Develop comprehensive and detailed automated evaluation systems for next-generation models to help assess model capability boundaries and inform research priorities.
  • Apply theoretical breakthroughs to real-world product challenges and advance the practical impact of AI applications.

Requirements:

  • Strong programming skills; proficient in Python and C/C++ in Linux environments. Familiar with PyTorch and mainstream frameworks for large model training and fine-tuning. Capable of independently building complex deep learning models and system modules, with solid debugging and performance optimization skills.
  • Experience with large-scale data preprocessing, generation, and augmentation. Understanding of data-driven model iteration workflows.
  • Familiarity with large model training pipelines, including distributed training, model parallelism, and training efficiency optimization.
  • Excellent problem-solving skills, a collaborative mindset, and effective communication abilities.

Preferred Qualifications：

  • Familiarity with high-performance computing frameworks such as CUDA, Triton, or Cutlass.
  • Experience with distributed reinforcement learning frameworks such as veRL, OpenRLHF, or Ray.
  • Knowledge of building large-scale RL environments involving browsers, code sandboxes, or simulated computer usage.
  • Experience with distributed training frameworks such as Megatron-Core or DeepSpeed, including multi-node training optimization and overlap efficiency of computation and communication.
  • Outstanding results in competitive programming contests (e.g., ACM/ICPC, NOI/IOI, Codeforces, Topcoder).
  • Contributions to well-known open-source large model projects or recognition in related competitions.
If you are interested in these job openings, please submit your resume and cover letter to shandahr@shanda.com. We also welcome assistance from recruitment agencies.