We build full-stack infrastructure that pushes generative media past the limits of today's hardware. We help developers and enterprises run image, video, and world models up to 100× faster and cheaper, at the same fidelity. Backed by Emergence Capital.
About the role
Nunchux makes visual generative models fast and efficient enough for production. As a Machine Learning Engineer on post-training and evaluation, you will build post-training pipelines that improve model efficiency and quality, measure the trade-offs, and turn the best recipes into reliable workflows for the models we ship. Develop post-training recipes: Establish and validate post-training recipes for image and video generation models.
What they're looking for
- Visual generative-model experience: Hands-on experience with image or video generation models, or with other multimodal visual systems. You understand the artifacts, failure modes, and quality trade-offs that matter in generated visual content
- Post-training or evaluation depth: Depth in one of two areas: post-training methods such as distillation or LoRA, or the evaluation of visual generative models
- ML engineering strength: Strong Python and PyTorch skills, with experience building post-training or evaluation code that others can run and maintain
- Training systems: Comfortable running and adapting post-training workloads across multiple GPUs with FSDP, DeepSpeed, or similar tools
- Experimental judgment: Able to design clean experiments, interpret the results, and make practical recommendations from the data
More about this role
Nunchux AI builds infrastructure that makes multimodal generative AI faster and cheaper to serve, and easier to build on. Founded by MIT PhDs Muyang Li, Yujun Lin, and Zhekai Zhang with CMU Professor Jun-Yan Zhu, Nunchux brings together deep research expertise and production systems experience. Our work is built on nearly a decade of research from MIT and CMU, including nunchaku project, whose models have surpassed 4 million downloads. We have top VC backing, and we build for enterprises and for millions of developers.
Nunchux makes visual generative models fast and efficient enough for production. As a Machine Learning Engineer on post-training and evaluation, you will build post-training pipelines that improve model efficiency and quality, measure the trade-offs, and turn the best recipes into reliable workflows for the models we ship.
Develop post-training recipes: Establish and validate post-training recipes for image and video generation models.
Build data pipelines: Curate and version the training and benchmark data used for post-training and model evaluation.
Build evaluation systems: Create benchmarks and automated judges, and run human preference studies, to measure...
Browse similar: AI jobs · AI startup jobs · Startup jobs · Remote jobs · San Francisco Bay Area