Download and run inference on your machine

1. Download Trained Model with Inference Code

After the training you can also the download the trained model and inference code, you can run this inference code anywhere.

  • Click on "Download Trained Model with Inference Code" and it will download the zip folder.

DLS Inference/Deploy page
  • Extract the zipped folder.

Downloaded zip and Extracted folder

Note :

  1. Install conda/Miniconda if not installed : Refer - https://docs.conda.io/projects/conda/en/latest/user-guide/install/windows.html

  2. When installing Miniconda3 Please check the “Add Miniconda 3 to my PATH environment variable” and install it (refer below screenshot)

  • Once the installation is completed, Open the ReadMe file from the extracted folder.

  • Follow each step in the ReadMe file.

  • Launch Command Prompt (Windows) or Terminal (Ubuntu) and then follow step 3 onward.

  • Before step 5, create a "test.csv" in the extracted folder with the two columns labeled Image and

    Digit label, respectively.

  • Add image paths to Image columns.

  • Also, add those images to the extracted folder, that you want to predict.

Images and test.csv added in the folder
  • So, after step 5 and 6, DLS will create "test_results.csv" inside the output folder and you will get the predicted value in the csv file.

  • You can also check your results on the command prompt .