AkôFlowHub
Explore
Back to Research
akoflow / machine-learning-model-conversion

Model conversion and quantization

Import into Desktop
Import version

Choose import version

akoflow::o/akoflow/machine-learning-model-conversion:v1

Export a learned affine model as real ONNX and TorchScript files, generate int8 and IEEE float16 coefficient representations in parallel, reload both executable formats, and compare their numeric outputs against the original model. The float16 branch runs on CPU and does not build or claim a TensorRT engine.

Public experiment
IMMUTABLE WORKFLOW VERSION

v1 / Official showcase: measured homeserver execution using a public immutable image

YAML

Workflow DAG

Loading workflow v1…

Code

Loading workflow…

Changes

Compare against

Choose compare against

Empty → v1

Loading changes…

SHA-256 8aeae316a21c3bf0bebb16dd538d8252c7c4483fde98efb9feed323f72d4814f