Steboss89

def matmul(mat1, mat2, mat3):
    r""" Function to multiply mat1 and mat2 
    returns mat3 
    Parameters
    ---------
    mat1: np.array, matrix A 
    mat2: np.array, matrix B 
    mat3: np.array, empty matrix C 
    
    Return 

Steboss89

import streamlit as st

# define a block where the tweet can be written
text = st.text_input("Insert tweet here")

# codify the input text in a compatible SageMaker json
input_df = pd.DataFrame({"text":text}, index=[0]).to_json(orient="split")

# here we are printing out results
st.subheader("Sentiment:")

Steboss89

import pandas as pd
import json
import boto3

# app name and region are global variables
global app_name
global region
app_name = 'NaiveBayesTest'
region = "eu-west-1"

Steboss89

import mlflow 
import os 

# PARSE YOUR MLFLOW INFO

# image build and push by mlflow
image_uri = "230178520806.dkr.ecr.eu-west-1.amazonaws.com/mlflow-pyfunc:1.28.0"
# the model uri
model_uri = "models:/SageMaker1/Staging"
# your region

Steboss89

import mlflow

run_id= "ade1040e1fd94a74b1da5ec0583fba63"
artifact_name = "model"
model_name = "NAME OF YOU MODEL"
mlflow.register_model(f"runs:/{run_id}/{artifact_name}", model_name)

Steboss89

#!/bin/bash

echo "Submit NB with CountVectorizer"
venv/bin/python multiple_models.py --run_name "naivebayes" --exp_name "sentiment_comparison" --model "naivebayes" --vectorizer "countvectorizer"

echo "Submit NB with TFIDF"
venv/bin/python multiple_models.py --run_name "naivebayes_tfidf" --exp_name "sentiment_comparison" --model "naivebayes" --vectorizer "tfidf"

echo "Submit LogReg with CountVectorizer"
venv/bin/python multiple_models.py --run_name "logreg" --exp_name "sentiment_comparison" --model "logreg" --vectorizer "countvectorizer"

Steboss89

import mlflow 
from mlflow.tracking.client import MlflowClient

# CLASS AND PIPELINE 

# MAIN 
# argparse ...

# set up the tracking and define the input arguments
mlflow_client = MlflowClient(tracking_uri=mlflow_tracking_uri)

Steboss89

import argparse

# MAIN
parser = argparse.ArgumentParser(description='Input arguments for MLflow testing in general')
parser.add_argument('--run_name', type=str, 
                    help='Name of the run within the experiment family')
parser.add_argument('--exp_name', type=str,
                    help='Name of the family experiment')
parser.add_argument('--model', type=str,
                    help='Model to be used: naivebayes, logreg, randomforest')

Steboss89

def training_process(model:str, 
                     vectorizer:str):
    r""" Function to create the training pipeline with cleaner and model 

    Parameters
    ----------
    model: str, type of model we want to run, see get_model function 
    vectorizer: str, type of vectorizer, `countvectorizer` or `tfidf`

    Return 

Steboss89

import mlflow 
# all imports 


# PREPROCESSING FUNCTION

# input dataframe
# clean dataframe 
# set up train and test
# count vectorizer