microsoft/phi-1

microsoft/phi-1 is a 1.3 billion parameter Transformer-based language model developed by Microsoft, specifically specialized for basic Python coding tasks. Trained on a curated dataset including Python code from The Stack v1.2, StackOverflow Q&A, code_contests, and synthetic Python textbooks, it achieves over 50% accuracy on the HumanEval benchmark for simple Python coding. This model is optimized for generating Python code from natural language prompts.