This is an advanced course on how to use the API ChatGPT to solve various tasks related to generating database, backend and data processing codes. It also covers limitations ChatGPT and how to improve its functionality by integrating with third-party APIs. The last part of the course expands on the use of open-source alternatives to implement similar on-premise models.
Each student will be assigned their own configuration of the virtual lab environment.
Additional details:
To participate in this course, you must have:
• PC/Laptop with internet access
• Updated web browser
This course is suitable for:
• Developers, programmers, system administrators who want to incorporate ChatGPT in their projects
• Developers, programmers who want to learn more about AI and natural language processing
• Professionals who want to learn advanced topics on how to use ChatGPT to improve their workflow and achieve more in less time
After completing this course, participants will acquire knowledge and skills such as:
● Use ChatGPT for specialized tasks like manipulating SQL data, defining API endpoints through Flask
● Integration ChatGPT in the scriptures
● Building applications around ChatGPT
● Using HuggingFace's pre-trained models.
● Leveraging hosted and local HuggingFace models.
Before attending this course, we recommend that participants:
● Be familiar with the concept of APIs and the basics of the Python programming language.
Module 1: ChatGPT Review
- Interacting with the interface
- Enhancing development experience
- Comparison to other models: BERT, GPT and T5
Module 2: ChatGPT for specialized tasks
- What tasks are best suited for ChatGPT automation?
- Creating a static HTML site
- Defining simple API endpoints through Flask
- Manipulating complex data through SQL
- Getting insights from Excel/CSV files through Pandas
Hands-on Lab
Module 3: ChatGPT API
- Calling ChatGPT from a Python script
- Handling Conversations
- Integrating ChatGPT in scripts
Hands-on Lab
Module 4: Building applications around ChatGPT
- ChatGPT limitations
- Compensating for limitations through external APIs
- Knowledge completion through the Wikipedia API
- Advanced mathematics through the WolframAlpha API
Hands-on Lab
Module 5: The need for an on-prem chatGPT: the HuggingFace ecosystem
- The problem with closed-source APIs
- Jackhammers vs toolboxes: why using the right tool matters
- Testing out a text/image generation model in the browser
Hands-on Lab
Module 6: Leveraging hosted and on-prem HuggingFace models
- Testing applications without the boilerplate
- Integrating an API endpoint in your application
- Moving a model to your own infrastructure
- Staying In the Loop: HuggingFace Spaces
Hands-on Lab
There are no recommendations at this time.
The course is not associated with any certification program.
