Linear 8-Bit Quantization

This section describes the following:

• Quantization Overview: Overview of linear quantization and considerations on model size and performance.
• Post-Training Quantization: Weight quantization of Core ML models using ct.optimize.coreml.linear_quantize_weights.
• Training-Time Quantization: Linear quantization of PyTorch models with data and fine-tuning, using ct.optimize.torch.quantization.LinearQuantizer.