quantization — Model Optimizer 0.27.1 (original) (raw)

modelopt.torch.quantization.calib

Calibrator classes.

modelopt.torch.quantization.compress(quant_model)

Compress model weights of quantized model.

modelopt.torch.quantization.config

This document lists the quantization formats supported by Model Optimizer and example quantization configs.

modelopt.torch.quantization.conversion

Quantization conversion/restore utilities.

modelopt.torch.quantization.export_onnx

Utility to export a quantized torch model to quantized ONNX.

modelopt.torch.quantization.extensions

Module to load C++ / CUDA extensions.

modelopt.torch.quantization.mode

This module contains the mode descriptor for the quantization mode.

modelopt.torch.quantization.model_calib

Calibration utilities.

modelopt.torch.quantization.model_quant

User-facing quantization API.

modelopt.torch.quantization.nn

Modules with quantization support.

modelopt.torch.quantization.optim

Deprecated.

modelopt.torch.quantization.plugins

Handles quantization plugins to correctly quantize third-party modules.

modelopt.torch.quantization.qtensor

Tensor Class for Real Quantization.

modelopt.torch.quantization.quant_modules

Deprecated.

modelopt.torch.quantization.tensor_quant

Basic tensor quantization functions.

modelopt.torch.quantization.triton

Triton quantization kernels.

modelopt.torch.quantization.utils

Quantization utilities.

Quantization package.