https://github.com/mahikshith/aws_sagemaker_ml_trainer

Using AWS -Sagemaker to train Machine learing model directly from jupyter notebook
https://github.com/mahikshith/aws_sagemaker_ml_trainer

Last synced: about 1 month ago
JSON representation

Using AWS -Sagemaker to train Machine learing model directly from jupyter notebook

• Host: GitHub
• URL: https://github.com/mahikshith/aws_sagemaker_ml_trainer
• Owner: mahikshith
• Created: 2024-12-09T11:17:36.000Z (about 2 months ago)
• Default Branch: main
• Last Pushed: 2024-12-09T11:24:54.000Z (about 2 months ago)
• Last Synced: 2024-12-09T12:24:30.931Z (about 2 months ago)
• Language: Jupyter Notebook
• Size: 145 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Directly connecting jupyter notebook to sagemaker for end-point inference

Using AWS -Sagemaker to train Machine learing model directly from jupyter notebook

### Most of the times we use AWS UI to access , open notebook inside sagemaker , train the model and inference from endpoint but what if we configure sagemaker using CLI and access it from jupyter notebook from your local machine

>Inshort we are just using AWS UI for creating IAM role , S3 bucket and the best part is we don't have to use sagemaker UI for training and inferencing from end-point

## Steps : 

### Configure AWS CLI & IAM role

- First create IAM role from ur AWS , set poicy access as admin or sagemaker access and create access key

- Download and Install Amazon CLI and configure

> aws configure

- paste the access & secret key

-Follow the code in the notebook

### Create S3 bucket & custom script

Notebook :  AWS_Sagenaker_mL_train.ipynb [https://github.com/mahikshith/AWS_Sagemaker_ML_trainer/blob/main/AWS_Sagenaker_mL_train.ipynb]

- Create S3 bucket and upload the data to S3 using the code in the notebook

- Write a custom script [sagemaker_script_ml.py] which contains code for sagemaker to train the model

- Create an entry point so that AWS sagemaker picks the code

- Create a new role for sagemaker service- execution and copy "ARN" and assign to role variable in the notebook

- Enter all the hyper-parameters for the endpoint and RUN the job

### Model -training and inferencing : 

- When we run the job , the model training starts under training-jobs

- Now create an endpoint for inferencing

- Stop or delete the end-point once the inference is done