AI & ML interests
LLM's
Recent Activity
View all activity
Organization Card
Celeris
We're a research lab building language models and inference systems for real-time intelligence. Our goal is to maximise useful intelligence per second.
Our work focuses on:
- Diffusion language models โ generating and refining text in parallel.
- Speculative decoding โ accelerating existing models with fast, accurate drafters.
- Decision models โ classification, routing, and structured decisions at millisecond latency.
- Inference Systems โ hardware aware architecture and optimised kernels.
Website ยท Documentation ยท API access
models 0
None public yet
datasets 0
None public yet