Data Science Portfolio

January 1, 2019 · 0 min · Greg Hilston

How To Locally Run Stable Difussion (Similar to Midjourney + DALLE-2)

If you just want to see examples of images I’ve generated, click here. Stable Diffusion is a new, free to use, model hosted by Hugging Face, which was produced by CompVis and Stability AI. This model takes in a text prompt, and produces an image. It was trained on 5.85 billion images. What’s most exciting about it, is it is freely available, and you’re able to run it on your own computer at home!...

September 1, 2022 · 3 min · Greg Hilston

Generating Images Using Diffusion Models (Similar To DALL-E 2)

FYI This post is much less useful, compared to my new post on Stable Diffusion. Reference This entire effort, and blog post, was inspired by a single Tweet by Simon Willison (@simonw). Hugging Face recently announced that new diffusion models have been released. For those that are unaware, diffusion models are able to take in some user input, some noise, and create content similar to what its been trained on....

August 9, 2022 · 2 min · Greg Hilston

Tmux

I’m always looking for more tools and ways to improve my productivity. One of my favorite tools that I’ve learned about is Tmux. I’m going to walk through the simple commands I use in Tmux every day and why I think Tmux is such an incredible tool. Put simply, Tmux is a terminal multiplexer. This means you can run a single Terminal application, but have multiple terminals controlled by a single screen....

October 20, 2021 · 6 min · Greg Hilston

Metaflow Open Source Contribution

March 30, 2021 · 0 min · Greg Hilston

Deep Learning with PyTorch: Optimizers

This is based on code from the following book The follow blog post walks through what PyTorch’s Optimzers are.e Link to Jupyter Notebook that this blog post was made from Pytorch comes with a module of optimizers. We can replace our vanilla gradient descent with many different ones without modifying a lot of code. %matplotlib inline import numpy as np import pandas as pd import seaborn as sns from matplotlib import pyplot import torch torch....

October 29, 2020 · 49 min · Greg Hilston

Tidy Tuesday

Tidy Tuesday is an excellent project that uploads a data set every week. There’s no real goal for the dataset, rather you’re asked to explore, play with and generally have a good time with the dataset. While the official repo calls for those to write their work in R, I personally use Python. Another good reference is David Robinson who does a live stream every week playing with the data for the first time live....

October 29, 2020 · 1 min · Greg Hilston

Deep Learning with PyTorch: Autograd

This is based on code from the following book The follow blog post walks through what PyTorch’s Autograd is. Link to Jupyter Notebook that this blog post was made from %matplotlib inline import numpy as np import torch torch.set_printoptions(edgeitems=2) Taking our input from the previous notebook and applying our scaling t_c = torch.tensor([0.5, 14.0, 15.0, 28.0, 11.0, 8.0, 3.0, -4.0, 6.0, 13.0, 21.0]) t_u = torch.tensor([35.7, 55.9, 58.2, 81.9, 56.3, 48....

October 28, 2020 · 3 min · Greg Hilston

Deep Learning with PyTorch: Parameter Estimation

This is based on code from the following book The follow blog post walks through what Parameter Estimation is. The goal here is to explain the theory. I have rewritten this notebook from the above book’s PyTorch Tensor implementation to be just in pure Numpy. Link to Jupyter Notebook that this blog post was made from Download FILE The story here is we’ll learn about Parameter Estimation by pretending we have two thermometers on our desk....

October 27, 2020 · 8 min · Greg Hilston

Data Science From Scratch To Production MVP Style: Deploy To AWS Lambda

Push Pipeline and Model To AWS S3 Create Buckets AWS S3 is a wonderful managed service that lets us upload files and access them later. We’ll leverage this by uploading our pickled files so our API can fetch them later. First we’ll create two “buckets”, think of these like a folder, one for our pipelines and one for our models. import boto3 session = boto3.Session(profile_name="personal") s3 = session.client("s3") s3.create_bucket(Bucket="data-science-from-scratch-pipeline") s3.create_bucket(Bucket="data-science-from-scratch-model") If you get any errors like:...

April 11, 2020 · 6 min · Greg Hilston