Update 11L default model, allow language to be used by more models
This commit is contained in:
@@ -55,9 +55,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
|
||||
|
||||
### Changed
|
||||
|
||||
- `auto_mode: True` for the `ElevenLabsTTSService`. This change
|
||||
significantly reduces the latency by disabling the chunk schedule and all
|
||||
buffers.
|
||||
- The default model for `ElevenLabsTTSService` is now `eleven_flash_v2_5`.
|
||||
|
||||
- `OpenAIRealtimeBetaLLMService` now takes a `model` parameter in the
|
||||
constructor.
|
||||
|
||||
Reference in New Issue
Block a user