- Date
TBC
- Time
09:30 – 17:30
(Total 8 hours)
- Language
Cantonese, Supplemented with English terminology
- Fee
HK$10,000 (Coming Soon)
Course Information
Course Name: Google Cloud Generative AI Fundamentals for Developer
Trainer: Zorro Cheng
Certificate: A minimum of 70% attendance rate is required for awarding of a completion certificate
Application Deadline: 7 days before the course
Remark: Please Bring your own laptop (BYOD) to classes.
What Will You Achieve
- Implementing a character-based text generator using an encoder-decoder neural network with TensorFlow Keras
- Using attention to improve the performance of machine learning tasks like machine translation and question answering
- Building end-to-end workflows for text classification and image captioning models with pre-trained BERT models
Who Is This Course For?
- Developers who want to build generative AI applications
Requirement
Be familiar with
- Application Development
- Systems Operations
- Linux Operating systems,
- Data Analytics/Machine Learning
Course Outline
- Learn how to implement a character-based text generator using an encoder-decoder neural network architecture with TensorFlow Keras.
- Understand how to preprocess text data and create input and target sequences for the neural network.
- Learn how temperature controls randomness vs determinism in decoding model outputs.
- See an example of generating creative text content like poetry using a trained encoder-decoder model.
- Learn how attention works
- How attention can be used to improve the performance of a variety of machine learning tasks, including machine translation, text summarization, and question answering.
- Learn how to load a pre-trained BERT NLP model from TensorFlow Hub and use it for text classification.
- Understand how to preprocess text data and fine-tune BERT for a downstream classification task like sentiment analysis.
- See how to export a trained BERT model and deploy it on Vertex AI to serve predictions through an online endpoint.
- Build an end-to-end workflow from loading BERT and training to deploying a text classification model on Vertex AI.
- Learn how to build an image captioning model by combining a CNN image encoder and RNN text decoder with attention.
- Understand how to preprocess image and text data, create training sequences, and tokenize text for the model.
- See how to train the encoder-decoder model end-to-end and generate captions by feeding images to the trained model.
- Learn how attention helps the decoder focus on relevant parts of the image when generating each word of the caption.
- Learn what Generative AI and Generative AI Studio are and how they allow you to generate content like text, images, and audio.
- Understand the core capabilities of Generative AI Studio including prompt design, conversation creation, and model tuning.
- See examples of using Generative AI Studio to design prompts, define conversation contexts, and launch tuning jobs to customize language models.
- Learn how Generative AI Studio provides an easy way to prototype and deploy generative AI models without coding through its graphical interface.
