Is OpenAI Whisper, Google TTS, or Piper TTS faster? ⏩

Is OpenAI Whisper, Google TTS, or Piper TTS faster? ⏩

I have to admit that I was wrong.

Switching ChatGPT versions wasn't the solution to my DIY Jarvis' latency issue. In the previous email, I revealed how I tested GPT-3 against GPT-4. Despite GPT-3 being faster, it didn't solve the issue because it was only one second faster. 

What exactly was the issue?

The issue was that I needed to shorten the delay between finishing my sentence while speaking to Jarvis and the start of Jarvis's response, which currently takes about 6 seconds.

Imagine you ask me a question, and then there's an awkward silence followed by a delayed response. This can really ruin the conversational experience. It's known as the response latency.

More debugging brought to light that the real issue was OpenAI Whisper.

OpenAI Whisper is an open-source automatic speech recognition system designed to transcribe speech accurately across various languages. 

In the context of Jarvis, it functions as the ears of the system. It lets Jarvis understand spoken commands and queries.

With OpenAI Whisper identified as the issue, what did I do next?

I ran more tests! This time, I compared OpenAI Whisper, Google TTS, and Piper TTS. 

Learn More About the Iron Man ChatGPT JARVIS course

Before running the tests, I expected the locally run Piper TTS to be the fastest. The results quickly showed me I was right, but the low quality of Piper TTS was still holding back Jarvis' performance. I knew there had to be a better way to solve this issue.

After lots of trial and error, what was the final solution to the latency issue?

I learned that using OpenAI Whisper was the best option because it gave the best results. And, I approached the latency issue using a technique called Sequential Sentence Chunking!

Sequential Sentence Chunking is a process where we break down text into smaller, manageable pieces to analyze it more easily. This technique allowed us to reduce the response latency from 6s to 1s! 

Curious about how I implemented this solution?

In our Iron Man ChatGPT Jarvis course, we’ll show you exactly how Sequential Sentence Chunking works and how to implement it in code, to improve your response latency 6x! 

The course shows you in detail every step needed to create your own Jarvis. The course includes the following modules:

• Module 1: Introduction and Project Overview - Explains the project's goals, necessary components, and potential applications.
• Module 2: Hardware Setup - Goes into how to assemble and connect the hardware components of the Jarvis Helmet.
• Module 3: Software Configuration - How to set up the Raspberry Pi, installing the necessary software, and writing scripts for voice recognition, chatbot functionality, helmet control, and system integration.
• Module 4: Future Potential and Conclusion - Dives into future improvements and gives resources to continue learning and share your project.

Unleash the Iron Man in you, and build your own DIY Jarvis for just $69 today.

Don't miss out. The special offer is only for today. Spots are filling up fast, and the price will increase tomorrow. 

🚨Prices are Set to Increase by $10 Tomorrow!!

Enroll in the Iron Man ChatGPT JARVIS course

Hold on, that's not all.  

If you have already enrolled in Augmented AI University, the Iron Man ChatGPT JARVIS course is already included in your subscription.

What is Augmented AI University all about?

Augmented AI University is a great resource to learn about everything related to artificial intelligence. 

Augmented AI University was created for anyone that is curious about AI, and wants to get hands-on with AI. 

Augmented AI University touches on everything from how AI is used in farming and the stock market to the latest in self-driving cars and recognizing objects.

Learn More About Augmented AI University

Consider the doors that a DIY version of Jarvis could open for you. The opportunities in the AI world are endless.

Here's a review by one of our students who have taken our other courses: