Introducing Speech Shrink, a powerful shortcut that leverages the capabilities of Private LLM to quickly summarize long speeches or presentations. By combining the convenience of Apple Shortcuts with the advanced language processing of Private LLM, Speech Shrink streamlines your workflow and saves you valuable time.

How Speech Shrink Works:

Highlight the entire transcript of the speech or presentation you want to summarize.
Activate Speech Shrink via the share sheet on iOS or system services on macOS.
Private LLM analyzes the transcript and generates a concise summary, capturing the key points and main ideas.
The summary is outputted, ready for you to use or share as needed.

Recommended Models and Settings:

For optimal performance, we recommend using the following models with Speech Shrink:

On Macs: Meta Llama 3 70B, Mixtral 8x7B Based Models, Solar 10.7B Based Models, or Meta Llama 3 8B Based Models
On iPhones: Meta Llama 3 8B Based Models

To fine-tune the summary generation, we suggest starting with a temperature of 0.7 and top-p sampling of 0.9. Adjust these settings based on your preferences for creativity and coherence.

Benefits of Using Speech Shrink with Private LLM:

Save time by quickly summarizing long speeches or presentations
Maintain privacy with local, offline processing of your data
Customize the shortcut to suit your specific needs
Integrate seamlessly with your existing workflow on iOS and macOS

Speech Shrink is just one example of how you can harness the power of prompt engineering, LLMs, and Apple Shortcuts to bring AI capabilities into your favorite apps. With Private LLM, you have the freedom to create custom workflows that enhance your productivity while keeping your data secure and private. Download Private LLM today and start exploring the possibilities of local AI.