The same bits everywhere
Two runs of an ordinary model on two machines rarely agree to the bit. A Laplace model is built so that they always do.
What it takes to answer the same way twice
- 1
Integer weights
When a model is compiled, its weights are quantized to integers. A run holds no floating point at all, so there is no rounding for two chips to do differently.
- 2
Integer operators
Every operator a model uses (a matrix product, a normalization, a softmax, a lookup) is written in integer arithmetic. The same inputs give the same outputs, to the bit.
- 3
A fixed order
A run is a fixed list of steps, each one a bounded slice of one operator. They are executed one after another, in the same order on every machine. Nothing is scheduled, batched or raced.
- 4
A seeded sampler
Where a model samples, it draws from a generator seeded by the seed on the receipt, so the same input and seed give the same output. A model that answers with one label, like the two on the network, takes the highest score, and the seed changes nothing.
- 5
One step function
The function that executes a step is one Rust crate, laplace-vm. It is compiled natively for nodes, to WASM for a browser tab and to SBF inside the program on Solana. One source decides a step everywhere.
One step, three builds
Reading the network