Get in Touch

Course Outline

Foundations of Edge AI and Nano Banana

  • Defining traits of edge-AI processing demands
  • Examining Nano Banana’s architecture and features
  • Contrasting edge versus cloud distribution approaches

Readying Models for Edge Implementation

  • Selecting models and establishing baseline metrics
  • Assessing dependencies and system compatibility
  • Preparing model exports for subsequent refinement

Strategies for Model Compression

  • Pruning methods and structural sparsity
  • Parameter reduction through weight sharing
  • Assessing the effect of compression on performance

Leveraging Quantization for Edge Efficiency

  • Techniques for post-training quantization
  • Workflows for quantization-aware training
  • Application of INT8, FP16, and mixed-precision strategies

Performance Acceleration via Nano Banana

  • Utilizing Nano Banana’s acceleration tools
  • Combining ONNX with hardware-specific backends
  • Testing the speed of accelerated inference

Releasing Models on Edge Hardware

  • Incorporating models into embedded or mobile software
  • Setting up runtime parameters and oversight
  • Resolving common deployment challenges

Performance Analysis and Trade-off Evaluation

  • Managing latency, data throughput, and heat generation
  • Balancing accuracy against operational speed
  • Applying iterative improvement methods

Guidelines for Sustaining Edge-AI Systems

  • Managing versions and ongoing upgrades
  • Handling model reversion and compatibility
  • Addressing security and data integrity factors

Conclusion and Future Directions

Requirements

  • A solid grasp of machine learning pipelines
  • Proficiency in developing models using Python
  • Working knowledge of neural network structures

Intended Learners

  • ML engineers
  • Data scientists
  • MLOps specialists
 14 Hours

Number of participants


Price per participant

Testimonials (1)

Upcoming Courses

Related Categories