Upload a JSONL file populated with a list of training examples.

Each line of the file should have an input and output key. The input should be an OpenAI chat completion, and the output should be formatted as an OpenAI assistant message.

Input

  • messages: Required. Formatted as a list of OpenAI chat completion messages.
  • function_call: Optional. Just like with OpenAI function calling, you can set this to { "name": "your_function_call_name" } to indicate that the model should be required to call the given function. For more information read OpenAI’s function calling docs.
  • functions: Optional. An array of functions available for the model to call. If you set the function_call argument, the function you asked it to call must be defined here. For more information read OpenAI’s function calling docs.

You can include other parameters from the OpenAI chat completion input format (eg. temperature), but they will be ignored since they aren’t relevant for training.

Output

Other Fields

  • split: Optional. One of “TRAIN” or “TEST”. If you don’t set this field we’ll automatically divide your inputs into train and test splits with a target ratio of 80:20.
...
{"input":{"messages":[{"role":"system","content":"You are a helpful assistant"},{"role":"user","content":"What is the capitol of Tasmania?"}],"functions":[{"name":"log_capitol","parameters":{"type":"object","properties":{"capitol":{"type":"string"}}}}]},"output":{"role":"assistant","content":null,"function_call":{"name":"log_capitol","arguments":"{\"capitol\":\"Hobart\"}"}}}
{"input":{"messages":[{"role":"system","content":"You are a helpful assistant"},{"role":"user","content":"What is the capitol of Sweden?"}],"functions":[{"name":"log_capitol","parameters":{"type":"object","properties":{"capitol":{"type":"string"}}}}]},"output":{"role":"assistant","content":null,"function_call":{"name":"log_capitol","arguments":"{\"capitol\":\"Stockholm\"}"}}}
...