具身世界模型推理INFRA工程师 - XiaomiRobotics
Tasks
- Accelerate embodied model inference
- Adapt multi token prediction for inference
- Collaborate with algorithm teams to improve inference speed
- Implement CFG parallelism adaptation
- Optimize expert parallelism for inference framework
- Optimize tensor parallelism for inference framework
- Quantize models using FP8
- Quantize models using NVF4
- Support model deployment and open source release
Perks/Benefits
- N/A
Skills/Tech-stack
CFG Parallelism | Diffusion Models | Expert parallelism | FP8 | Multi Token Prediction | Multimodal AI | NVF4 | Tensor Parallelism
Education
Related jobs
-
机器人VLA算法研究员 - XiaomiRobotics CNY 500K-500KDeep learning | Diffusion Models | Language Modeling | Machine Learning | Mixture of ExpertsEntry-level Full Time北京21h ago
-
VLA训练infra算法工程师 - XiaomiRobotics CNY 240K-480KBF16 | C++ | CPU/memory optimization | CUDA | Data pipelineMid-level Full Time北京21h ago
-
多模态大模型算法工程师（Vlm / 自动驾驶方向） CNY 180K-264KAgent modeling | Autonomous Driving | Autoregressive modeling | BEV | Behavior ModelingEntry-level Full Time北京、苏州3d ago
-
Entry-level Internship北京9d ago
-
硬件Ai学术追踪实习生 CNY 25K-37KAdaptive Noise Cancellation | Antenna design | Beamforming | Bluetooth | CSTEntry-level Internship北京9d ago
-
Senior Software Engineer, 3D/4D Reconstruction CNY 417K-540K3D Computer Vision | Autonomous Driving | Computer Vision | Deep learning | Dense ReconstructionSenior-level Full TimeChina, Beijing11d ago
-
Computer Vision | CoreML | Deep learning | Diffusion Models | GLSLSenior-level Full TimeBeijing, China17d ago
-
Ai工程师 CNY 180K-300KAI infrastructure | AIGC | Automation | Deep learning | Language ModelsCollaborative culture | Growth opportunities | Latest hardware access | Technical community impactMid-level Full TimeShenzhen24d ago