Responsibilities:

1. Develop distributed artificial intelligence systems and deploy them on large-scale clusters or clouds.
2. Develop and optimize algorithm systems from specific scenarios and problems, and produce solutions applied to the scenarios.
3. Participate in the integration design and optimization of artificial intelligence technology and existing tools to improve product performance.
4. Write high-quality scientific papers and have the opportunity to serve as the first author of important papers.

Requirements:

1. Proficient in TensorFlow/PyTorch, Ray/DeepSpeed/NVIDIA Megatron, and familiar with the internal operating mechanisms of these systems.
2. Familiar with various optimization algorithms and model architectures, proficient in Python or C++ optimization algorithm libraries, including various classical algorithms and models based on gradients (BERT, GPT-3, Swin Transformer, ViT, MLP-Mixer).
3. Familiar with SaaS, architecture, compilers, networks, CUDA, and other related knowledge or have relevant project experience.
4. Strong programming and engineering implementation abilities. Those who have won programming competition awards or published high-quality papers will be given priority in recruitment.
5. Bachelor's degree, majoring in computer science, software engineering, electronic information, automation, mathematics, physics, or other artificial intelligence-related majors.