Releases: toshiakit/MatGPT
MATLAB Vault and Focus Mode Support
What's Changed
- Updated README, Added MATLAB Vault and Focus Mode Support by @toshiakit in #41
Full Changelog: 2.0.5...2.0.6
Move 'LLMs for MATLAB' to submodule
MatGPT now pulls in the LLMs for MATLAB as a submodule to better track developments with this dependency.
Support for Whisper API and GPT-4o Model
MatGPT supports voice chat via Whisper API and the new GPT-4o "omni" model.
Support the latest models released by OpenAI on April 9, 2024
OpenAI released the updated models gpt-4-turbo and gpt-4-turbo-2024-04-09, superseding the preview models.
LLMs with MATLAB was updated to reflect this change, and this update will make MatGPT current with the latest release of LLMs with MATLAB.
Also, the DALL-E 3 image generation is now handled via function calling using gpt-3.5-turbo, to enable chat experience in image generation.
Full support for GPT-4 vision and DALL·E 3 image generation
Merge pull request #29 from toshiakit/dev Adding support for GPT-4 Vision & DALL·E Image Generation
Streaming Support
MatGPT now supports the streaming API from OpenAI where response from LLMs are displayed as they come in. The app has been updated to include clickable follow-up prompts to further explore a topic.
New LLM framework and prompt templates
MatGPT is now based on the official MATLAB library for interfacing with LLMs including models from OpenAI.
Added additional support for reading and chatting with web links and files.
GPT-4 and new OpenAI model support
- Supports GPT-4 and new models from OpenAI
- Updated token usage
GPT-4 and new OpenAI model support
Merge pull request #17 from toshiakit/dev GTP-4 and new model support
More robust table conversion
Fix the bug that fails to detect tables when the table header separator doesn't include spaces.