Over the past year, we have built a number of fine-tuned models for developers, including classification, embedding, retrieval and summarization/generation for source, documents, discussion and tasks/issues.
Interested to learn what the HN community thinks, as well as the clients/features we built around them.