Release 0.7.9

Released 2018-12-13.

• Added per-model rate limits
• Fixed token counting for long prompts
• Improved streaming reliability

Release 51682 · Release 51683 · Release 73946

Back to the changelog