2 ms·
We've been using `gpt-4-1106-vision-preview` and simply prompting the model to return json, with excellent results: https://github.com/OpenAdaptAI/OpenAdapt/pul
by abrichr 3y ago
We've been using `gpt-4-1106-vision-preview` and simply prompting the model to return json, with excellent results: https://github.com/OpenAdaptAI/OpenAdapt/pull/610 https://github.com/OpenAdaptAI/OpenAdapt/pull/610 (work in progress).