Commit sample inputs to protocol circuits #10602
Assignees
Comments
|
@LeilaWang I'm not sure on the best way of generating a set of correct inputs. Could someone from the protocol circuits team help generate these please? |
TomAFrench
pushed a commit
that referenced
this issue
Dec 11, 2024
Closes #10602 Generated by running `fs.writeFileSync("./SampleInputs.toml", TOML.stringify({inputs: mapped }))` on the circuit inputs (`mapped`) in `yarn-project/noir-protocol-circuits-types/src/index.ts` on the relevant conversion fns. Any test that involves the protocol circuits will reach those. I chose `yarn-project/prover-client/src/orchestrator/orchestrator_multiple_blocks.test.ts` simply because it's fast and uses all rollup circuits. @TomAFrench if this is what you were looking for and you also need kernel circuits, I'm happy to add those. Note: I named them `SampleInputs.toml` rather than `Prover.toml` because the latter is in `gitignore`.
lucasxia01
pushed a commit
that referenced
this issue
Dec 11, 2024
Closes #10602 Generated by running `fs.writeFileSync("./SampleInputs.toml", TOML.stringify({inputs: mapped }))` on the circuit inputs (`mapped`) in `yarn-project/noir-protocol-circuits-types/src/index.ts` on the relevant conversion fns. Any test that involves the protocol circuits will reach those. I chose `yarn-project/prover-client/src/orchestrator/orchestrator_multiple_blocks.test.ts` simply because it's fast and uses all rollup circuits. @TomAFrench if this is what you were looking for and you also need kernel circuits, I'm happy to add those. Note: I named them `SampleInputs.toml` rather than `Prover.toml` because the latter is in `gitignore`.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
There's currently no way for the noir team to benchmark execution time of the protocol circuits as we don't have any sample inputs that will successfully execute. It would be very helpful if a set of sample inputs were committed alongside the circuits in
Prover.tomlfiles.This would be very useful in preventing execution time regressions.
The text was updated successfully, but these errors were encountered: