Releases: nerve-sparks/iris_android
Releases Β· nerve-sparks/iris_android
Version 0.3 Beta
Release Log: [Version 0.3]
π New Features
-
New and Robust UI #72
- A completely redesigned user interface for an intuitive and seamless experience.
- Enhanced responsiveness and accessibility for all devices.
-
Hugging Face Models Integration #72
- Added support for downloading and using models directly from Hugging Face.
- Streamlined model import process for faster and efficient usage.
-
Efficient Model Management #72
- Improved tools to manage and switch between models effortlessly.
- Added functionalities to view model details and manage storage efficiently.
-
Speech-to-Text and Text-to-Speech
- Introduced Speech-to-Text functionality to transcribe input directly.
- Added Text-to-Speech feature to convert responses into natural-sounding audio output.
-
Hyperparameter Tuning #77
- Added controls for fine-tuning hyperparameters:
- Temperature, Top-p, Top-k, N-threads.
- Customize your model's behavior and performance with ease.
- Added controls for fine-tuning hyperparameters:
-
Benchmarking
- Measure model performance with detailed benchmarking tools.
- Includes inference speed, memory usage, and accuracy metrics.
-
Applies Chat Template Internally #50
- Automatic application of chat templates to improve conversational context and flow.
π Bug Fixes and Improvements
-
Inference Stability
- Fixed issues causing instability during model inferencing.
- Improved reliability for long-running sessions.
-
UI Glitches and Scrolling
- Resolved glitches in the UI that occurred during scrolling.
- Enhanced smoothness and performance of scrolling animations.
π Whatβs Next?
Stay tuned for more exciting features, enhanced model support, and further optimizations in the next update!
We welcome your feedback and contributions. π
Version 0.2 Beta
Version 0.2 Beta
Pre-release
Pre-release
Version 0.2 Beta
Change Log:
Version 0.1 Beta
First beta release of Iris LLM chat app
Install only from file manager.
Download model only on Wi-Fi.
Start asking questions.