Enable output parser for models not support json schema.

Eladlev / AutoPrompt

A framework for prompt tuning using Intent-based Prompt Calibration

Apache License 2.0

1.86k stars 149 forks source link

Enable output parser for models not support json schema. #56

Open vincilee2 opened 3 months ago

vincilee2 commented 3 months ago

Currently, not all output_schemas support customized output parsers.

For example in the classification output schemes, only json schemas are available, meaning this prompt only works well for the models which support json schema output.

To be compatible with other LLMs, a customized output parser is required.

Eladlev commented 3 months ago

We also support the 'completion' mode (for non-OpenAI models), this can be done by changing the meta prompts to meta_prompts_completion

Regarding supporting more methods to enforce the structure, we would like to integrate libraries such as LM Format Enforcer