Clarify GPT-4 model

[rodion-m]

Please clarify what GPT-4 models us used in the benchmark (4o or Turbo) and also, please add them both into the benchmark.

[booydar]

Thanks for pointing out this lack of clarity, I'll update the readme and paper text. For main experiments we used the gpt-4-0125-preview, and for retrieval-augmented generation we employed the GPT-4-turbo version. Evaluation results for each model can be found in the predictions_06_2024 branch.