Open tusharmath opened 2 weeks ago
Currently autogenerated configurations can sometimes have not-so-readable type names. We want to use LLMs to produce more meaningful type-names
Before
type T1 { name: String age: Int gender: Gender } type T2 { name: String address: Address reviews: [Review] }
After A simple instruction like below to chat gpt can produce a more meaningful configuration
Generate better names for T1 and T2 and produce a new schema file with the new names.
type Person { name: String age: Int gender: Gender } type Business { name: String address: Address reviews: [Review] }
Technical Requirements
Additional Information A sample transformer is written over here https://github.com/tailcallhq/tailcall/tree/main/src/core/config/transformer
Currently autogenerated configurations can sometimes have not-so-readable type names. We want to use LLMs to produce more meaningful type-names
Before
After A simple instruction like below to chat gpt can produce a more meaningful configuration
Technical Requirements
Additional Information A sample transformer is written over here https://github.com/tailcallhq/tailcall/tree/main/src/core/config/transformer