monk_v1

Website

Monk is a low code Deep Learning tool and a unified wrapper for Computer Vision.

   

Documentation

• Introduction
• Setup
• Quick Mode
• Update Mode
• Expert Mode
• Hyper Parameter Analyser
• Compare Experiments
• Resume Training

Create an image classification experiment.

• Load foldered dataset
• Set number of epochs
• Run training

ptf = prototype(verbose=1)
ptf.Prototype("sample-project-1", "sample-experiment-1")
ptf.Default(dataset_path="./dataset_cats_dogs_train/", 
                model_name="resnet18", freeze_base_network=True, num_epochs=2)
ptf.Train()

Inference

img_name = "./monk/datasets/test/0.jpg";
predictions = ptf.Infer(img_name=img_name, return_raw=True);
print(predictions)

Compare Experiments

• Add created experiments with different hyperparameters
• Generate comparison plots

ctf = compare(verbose=1);
ctf.Comparison("Sample-Comparison-1");
ctf.Add_Experiment("sample-project-1", "sample-experiment-1");
ctf.Add_Experiment("sample-project-1", "sample-experiment-2");
    .
    . 
    .
ctf.Generate_Statistics();

RoadMap

• [] Incorporate pep coding standards
• [] Functional Documentation
• [] Tackle Multiple versions of pytorch, keras, gluon
• [] Standardize folder structure for next feature additions - object detection, image segmentation
• [] Add support for tensorflow-2.0
• [] Add unit-testing

Copyright

Copyright 2019 onwards, Tessellate Imaging Private Limited Licensed under the Apache License, Version 2.0 (the "License"); you may not use this project's files except in compliance with the License. A copy of the License is provided in the LICENSE file in this repository.