Would it be possible to release the evaluation scripts of baselines for CodeSage including CodeBert, Starencoder and so on? It seems that the evaluation script for CodeSage cannot be directly used to evaluate baseline models.
For the baseline models, you can take a look at the CodeBERT repository which is the basis of our code for finetuning CodeSage and evaluation. We do not plan to release evaluation scripts for the baselines.
Would it be possible to release the evaluation scripts of baselines for CodeSage including CodeBert, Starencoder and so on? It seems that the evaluation script for CodeSage cannot be directly used to evaluate baseline models.