Automating Code Review Activities 2.0 (datasets, models and results)

Resources related by the research work "Automating Code Review Activities 2.0".

• automating_code_review.zip contains the material to successfully run our Colab notebooks;
• dataset.zip contains all the preprocessed datasets used in our work;
• generate_prediction.zip contains the material to successfully generate predictions using a T5 model chekpoint;
• models.zip contains the (best) checkpoints of the fine-tuned T5 models;
• results.zip contains our results;
• tokenizer.zip contains the Sentencepiece model and vocabulary trained on our pre-training dataset.

More information in the replication package of our work: code_review_autmoation