The Token Company
Member of Technical Staff, Research
San Francisco. Posted 5 months ago
The role
You will own the model training stack end to end including data, architecture, training, evaluation, and shipping compression models into production. You will design and train models on NVIDIA B200s.
What you will do
- Own the model training stack end to end
- Design and train models on NVIDIA B200s
- Handle data, architecture, training, and evaluation
- Ship compression models into production
Required qualifications
- Trained models from scratch
- Owned data, architecture and a training loop
- Strong in ML fundamentals
- Fluent with transformers
- Comfortable turning a paper or a rough idea into a real training run
- Learn fast
- Care about production
Nice to have
- Pretrained a transformer, or done serious post-training or RL on one
- Built a novel architecture or training method with results to back it up
- Shipped a model you trained into something people actually use
Benefits
- Competitive base salary
- Significant equity
- Cover housing
- Cover food
- Cover laundry
- Cover cleaning
- Sponsor visas
About the company
The Token Company is a seed stage startup in San Francisco. They train machine learning models that compress raw LLM inputs to cut inference costs.