Build experimental dataset with synthetic prompts for EVMind

[brunneis]

Build an experimental dataset with generated prompts given Solidity code with different levels of detail.

• ~300k prompts for ~100k code samples from seyyedaliayati/solidity-dataset.
• For each code sample, three prompts were generated for three profiles: expert, average user, and beginner.

[brunneis]

https://github.com/EveripediaNetwork/iq-code-evmind/blob/master/code-to-prompt/seyyedaliayati-solidity-dataset-300k-prompts.txt