v1.81.0-patch4
📦 litellmView on GitHub →
🐛 3 fixes🔧 3 symbols
Summary
This patch release focuses on stability, fixing streaming issues with Azure OpenAI and correcting token usage calculation for Gemini 1.5 Flash.
🐛 Bug Fixes
- Fixed an issue where the `completion` endpoint was incorrectly returning an error when using `azure_openai` with `stream=True`.
- Resolved a bug causing incorrect token usage calculation for `gemini-1.5-flash` when using function calling.
- Fixed an issue where `anthropic` models were failing when `max_tokens` was set to 0.