​
Introduction

Inferless allows you to deploy your model using CLI, which has similar file structure requirements to the GUI platform.

​
Mandatory File

When you want to deploy your models through CLI, you must have the file app.py. When you run the command inferless init, you are required to have this file in your root folder. app.py: This file will have the InferlessPythonModel class, which will contain the below functions(Mandatory)
  1. def initialize(self)
  2. def infer(self, inputs)
  3. def finalize(self)
input_schema.py This file should have all the inputs that you are taken 
INPUT_SCHEMA = {
    "prompt": {
        'datatype': 'STRING',
        'required': True,
        'shape': [1],
        'example': ["There is a fine house in the forest"]
    }
}
While running the command inferless init, all the other files required for the deployment will be created and the user can update these files according to their requirements.

​
Files structure

This is an example of the file structure required for the deployment of the CodeLlama-34b-python model using Inferless-CLI. 
codellama-34b-python/
├── app.py
├── inferless-config.yaml
├── inferless.yaml
└── input_schema.py 
# Depreciated 
├── input.json
└── output.json
User can update the following file according to their requirements.
  • inferless-runtime-config.yamlThis file will have all the software packages and the Python packages required for the model inferencing.
  • inferless.yamlThis file will have all the configurations required for the deployment. Users can update this file according to their requirements.
  • input_schema.py This file will have all the configurations required for the deployment. Users can update this file according to their requirements.
Deprecated - Input / Output Json
  • input.jsonThis file will have the key for the input parameter. Whenever you change the name of the key in the app.py, update it accordingly.
  • output.jsonThis file will have the name of the output key that the def infer the function is going to return.
Once these files are ready, you can use the command inferless deploy to deploy your model.