5 ms·
The embeddings on their own are insufficient for this. You need some kind of sequential model to generate the code. It should be possible to build your own mod
by nl 4y ago
The embeddings on their own are insufficient for this. You need some kind of sequential model to generate the code.
It should be possible to build your own model to do this instead of GPT-4 if you are so inclined. I don't know how the quality would compare but there are various specialized code-specific models already around (and more coming) that work quite well.
- yao420 4y agoWhat are the best ones for code? I’m looking to write and measure the coverage of generated fuzzers and unit tests.
- nl 4y agoI think the Salesforce ones[1] and the CarperAI[2] one. BigCode from HuggingFace will be coming soon too. [1] https://huggingface.co/Salesforce https://huggingface.co/Salesforce and expand "models" codegen-mono is Python, codegen-multi is multi language I think. [2] https://carper.ai/diff-models-a-new-way-to-edit-code/ https://carper.ai/diff-models-a-new-way-to-edit-code/