Multigen

Multigen requests a response from the AI in parts, the first request is 50 tokens, the rest are 30. Requests are made until the requested number of tokens is reached, or stop tokens are received.

*It is not used in OpenAI because OpenAI can stop generating with stop tokens