[Update] Azure's Memory Leak Crisis

Hi all! So I noticed there was a ton of hype in a Discord server, so I went to check if Azure had done anything, and they just dropped a huge update.<br><br>> **Optimization for Latency and Token Usage**: With the launch of Azure OpenAI 2023-05-10, you can now specify the desired response latency and token usage when calling the models through the API. This update allows you to optimize your usage for performance and cost. This update applies to both the Davinci and Bard models.<br><br>This means that we can now set the API to:<br>- 1.5x, 2x, 3x, or 5x times faster latency<br>- 20%, 30%, 40%, 50%, or 67% less tokens<br><br>You can set this per request, so you can high latency for less tokens when just answering easy questions, and higher latency with more tokens when generating.