donaldsebleung/california-housing-linear-simple

Predicting California house prices with MindSpore

A simple linear regression model to predict house prices in California

Overview

A simple linear regression model to predict house prices in California.

This model was trained on the California Housing Dataset using MindSpore 2.8.0 and exported in MindIR format as california-housing-linear-simple.mindir.

Dependencies

• MindSpore 2.8.0
• scikit-learn 1.8.0

The dependencies are specified in the included requirements.txt.

pip install -r requirements.txt

Usage

Load the model weights and initialize the mindspore.nn.GraphCell.

import mindspore
import mindspore.nn as nn

graph = mindspore.load('california-housing-linear-simple.mindir')
model = nn.GraphCell(graph)

Model inputs

A mindspore.Tensor of shape (n, 8) where n is the number of input samples.

Each input feature has dtype=mindspore.dtype.float16.

Model outputs

A mindspore.Tensor of shape (n, 1) where n is the number of output predictions.

Each output prediction has dtype=mindspore.dtype.float16.

End-to-end example

See the included notebook 01-load-infer.ipynb for details.