 |
ML Newsnew | past | best | rss | submit | about |
| Evaluating Large Language Models Trained on Code (paper about GH copilot model) (arxiv.org) | |
| 4 points by homarp on Jul 8, 2021 | hide | past | pdf | 1 comment on HN | |
|
In plain words: A language model trained on public GitHub code turns plain-English descriptions into Python programs. On a new set of coding problems it solved 28.8%, versus none for the same model without code training; drawing many guesses and testing them got most problems right. Abstract · Evaluating Large Language Models Trained on CodeWe introduce Codex, a GPT language model fine-tuned on publicly available code from GitHub, and study its Python code-writing capabilities. A distinct production version of Codex powers GitHub Copilot. On HumanEval, a new evaluation set we release to measure functional correctness for synthesizing programs from docstrings, our model solves 28.8% of the problems, while GPT-3 solves 0% and GPT-J solves 11.4%. Furthermore, we find that repeated sampling from the model is a surprisingly effective strategy for producing working solutions to difficult prompts. Using this method, we solve 70.2% of our problems with 100 samples per problem. Careful investigation of our model reveals its limitations, including difficulty with docstrings describing long chains of operations and with binding operations to variables. Finally, we discuss the potential broader impacts of deploying powerful code generation technologies, covering safety, security, and economics. | |
| add comment on HN |
A distinct production version of Codex powers GitHub Copilot.
On HumanEval, a new evaluation set we release to measure functional correctness for synthesizing programs from docstrings, our model solves 28.8% of the problems, while GPT-3 solves 0% and GPT-J solves 11.4%.
Furthermore, we find that repeated sampling from the model is a surprisingly effective strategy for producing working solutions to difficult prompts. Using this method, we solve 70.2% of our problems with 100 samples per problem.
Careful investigation of our model reveals its limitations, including difficulty with docstrings describing long chains of operations and with binding operations to variables. Finally, we discuss the potential broader impacts of deploying powerful code generation technologies, covering safety, security, and economics.
HumanEval: https://github.com/openai/human-eval