Can't generate proof with larger input

Noir
1

Hi, I’m trying to generate proof with Noir JS for a linear regression circuit, but the issue is for same circuit (with dataset size changed) with fewer data size, the proof is generated and verified correctly, but when I increase the data length I encounter this error.

Error [RuntimeError]: null function or function signature mismatch

image
image1256×656 218 KB

But if I try running nargo prove command with same inputs, the proof is generated correctly.

Here is my nargo info output:

image
image891×129 12 KB

and my nargo --version:
image

Thanks for the help.

2

Can you share more info about your circuit source code and the inputs you are using?

3

Since this is part of an ongoing research project, I am unable to share the exact source code. However, I can provide an overview of the circuit. It’s a 2D linear regression model where the inputs are two arrays, X (independent variable) and Y (dependent variable), both of a fixed length and of type i64.

4

a colleague of mine faced a similar issue. its likely a wasm error when the constraint size is too large (probably a wasm limit error throwing something else).

Do you think you would be able to rewrite the circuit to use recursion? Recursive Proofs | Noir Documentation