Hi, I have tried your product. It is awesome and amazing! Did you make it by finetuning an open-source llm?
I am a student from China and I want to make an individual generative text2sql model to facilitate the team's daily work. But our model will still have many problems like after 2 months' training, for example, When we try to improve the model's ability by supplementing the corpus with the missing ability, it forgets the other skills as it learns the task. could you give us some advice?
