[Weave] Extractable JSON from Prompt responses
When a prompt produces an output, there can be a step or automated action to extract the keys into a workflow variable
When a prompt produces an output, there can be a step or automated action to extract the keys into a workflow variable
Log in to comment and vote
Comments1
Shuwei Li
Jan 30, 2024
Found some interesting article. To ensure the return_result always in JSON format:
https://www.lamini.ai/blog/guarantee-valid-json-output-with-lamini#custom-parsers
Important information from the article:
Try our API to guarantee valid JSON output now: https://lamini-ai.github.io/rest_api/completions/
[1] JSONformer: https://github.com/1rgs/jsonformer
[2] OpenAI JSON Mode: https://platform.openai.com/docs/guides/text-generation/json-mode