High performance Inference.
Any model. Any hardware.
No compromise.
Built to redefine performance, flexibility, and portability_
>
Develop in zig: A modern, fast, and reliable language.
Why ZML?_
>
Very high performance.
>
No Python.
>
Zero code change between:
NVIDIA CUDA, AMD ROCm,
Google TPU, AWS Neuron.
>
Flat latencies.
