ammarnasr/the-stack-java-clean

Dataset 1: TheStack - Java - Cleaned Description: This dataset is drawn from TheStack Corpus, an open-source code dataset with over 3TB of GitHub data covering 48 programming languages. We selected a small portion of this dataset to optimize smaller language models for Java, a popular statically typed language. Target Language: Java Dataset Size: Training: 900,000 files Validation: 50,000 files Test: 50,000 files Preprocessing: Selected Java as the target language due to its… See the full description on the dataset page: https://huggingface.co/datasets/ammarnasr/the-stack-java-clean.