Detailed Accuracy Metrics for Individual Benchmarks

[taesiri]

Hi,

First of tall thank you so much for releasing this awesome model.

We want to include Snowflake-Arctic in our study, and we need to have detailed accuracy metrics for individual benchmarks. The official article syas that the "Common Sense" metric is an average of 11 metrics, and the "Coding" category also appears to be an average of HumanEval and MBPP. Could you provide the detailed scores for these individual benchmarks?

Thanks!