amd-quark

library0.11.2pypypi✓ verified 80d ago

AMD Quark is a comprehensive cross-platform toolkit for quantizing deep learning models, supporting PyTorch and ONNX. It enables model optimization for various hardware backends with minimal accuracy loss. Current version: 0.11.2. Release cadence is irregular.

ai-ml

pip install amd-quark

INSTALL

IMPORT

SIG · AMD-QUARK

amd-quark

ai-mlpythonv0.11.2

harness data pending

Install & Compatibility

Where this runs

No compatibility data collected yet for this library.

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

version

✓ from quark import version

✗ from quark.quantization import quantize

quantize

✓ from quark.quantization import quantize

Basic quantization of a PyTorch model.

import torch from quark.quantization import quantize from quark.quantization.api import Config model = torch.nn.Linear(10, 5) config = Config(calibration_method='percentile', calibration_percentile=99.9) q_model = quantize(model, config) print(q_model)

Debug

Known issues

breakingPython 3.12 is not supported (requires <3.13,>=3.9.0). Install Python 3.9-3.11.

fix

Use Python 3.9, 3.10, or 3.11.

affects: 0.11.2

deprecatedThe old import `from quark import quantize` has been removed. Use `from quark.quantization import quantize`.

fix

Update imports to use quark.quantization submodule.

affects: >=0.10.0

gotchaCalibration method names are case-sensitive and must match exactly (e.g., 'percentile', not 'Percentile').

fix

Use lowercase calibration method names.

affects: 0.11.x

Upgrade

Version history

0.11.2latest on PyPI · released May 4, 2026

Audit

Dependencies

No dependency data recorded yet.

Agent activity

23 hits · last 30 days

node

OpenAI (training)

Resources

No resource links recorded.