Release 0.34.1

Released 2022-01-22.

• Added per-model rate limits
• Fixed token counting for long prompts
• Reduced idle memory use

Release 99398 · Release 99399 · Release 29941

Back to the changelog