Fix trusted setup outputs by using correct transcript.json

[CarlBeek]

Problem

#2 used an incorrect version of the transcript.json resulting in incorrect output_setups.

Solution

Re-run with the final transcript sha256: 8ed1c73857e77ae98ea23e36cdcf828ccbf32b423fddc7480de6

Detailed explanation

Due to me working too quickly, and not verifying which transcript I used to create the files in #2, I ended up using the incorrect version of the transcript which resulted the output_setups being entirely incorrect. This happened due to me using a random copy of the transcript.json I had lying around, instead of re-downloading it, and then I did not verify the file was actually correct.

The updated output_setups in this PR are from transcript sha256 hash 8ed1c73857e77ae98ea23e36cdcf828ccbf32b423fddc7480de658f9d116c848 which yields the same g1_lagrange and g2_monomial outputs as before #2.

Attribution

Huge thank you to @zilm13 for finding this! 🙏🙏

This was a stupid mistake on my part and this is a great example of why open source is so important. Trust, but verify!

[jtraglia]

LGTM 👍 I've checked the following:

• The old reference tests pass with the new 4096 trusted setup.
• The old values (g1_lagrange, g2_monomial) haven't changed for all files.
• The number of values in g1_monomial is correct for all files.
• The g1_monomial values for the 4096 file match what c-kzg produces.

[CarlBeek]

Thank you for checking my work @jtraglia!