Week 2

MC - Sep 4 - - Dev Community

This week marked a significant shift in my project as I moved from learning about generative AI to conducting practical experiments using Python. The first few days were dedicated to exploring how text-based AI models respond to different prompts. These experiments helped me understand the nuances of AI-generated content and how to fine-tune outputs for specific needs.

The highlight of the week was the successful setup and initial use of the Whisper model, a speech-to-text AI. Starting on Day 12, I began experimenting with voice input, using Python to transcribe speech into text. These experiments were successful, showing that Whisper could accurately handle various speech inputs. However, I observed that the model’s output was more formal and written in style, which may not be ideal for my project’s needs. As a result, I plan to begin fine-tuning Whisper next week to produce more natural and conversational language, better suited to the Elderly Care System.

Overall, this week was highly productive, with significant progress made in both understanding and applying generative AI. The experiments with Whisper, in particular, have laid a solid foundation for further development, particularly in enhancing the accessibility and functionality of the Elderly Care System through voice input capabilities.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Terabox Video Player