Speaker 1: Jeremy Howard
0:00 | Welcome and Introduction 0:53 | Course Materials and Resources 1:46 | Chapter Quizzes and AIquizzes.com 2:39 | Course Forums and Summarizing Topics 4:14 | Show Us What You’ve Made Post 6:38 | Putting a Model in Production 7:16 | Data Cleaning and Chapter 2 of the Book 8:09 | Jupyter Notebook Extensions and Navigation 9:48 | Replacing Bing with DDG for Image Search 10:56 | Grizzly Bear Detector and Python/Fast.ai Help 12:47 | Training a Model Before Cleaning Data 13:11 | Data Block, ShowBatch, and Resizing Images 14:52 | Random Resized Crop and Data Augmentation 16:57 | Data Augmentation and Image Copying 18:07 | Training the Model and Confusion Matrix 19:09 | Confusion Matrix and Classification Interpretation 20:33 | Plot Top Losses and Loss Measurement 22:08 | FastAI Image Classifier Cleaner 23:40 | Cleaning Training and Validation Sets 25:24 | Data Cleaning and GPU Memory Management 26:19 | Asking Questions and Watching the Video 27:18 | Putting the Model in Production with Hugging Face Spaces 27:38 | Gradio and Tanishq Abraham’s Blog Post 28:08 | Tanishq Abraham and the Fast.ai Community 29:00 | Creating a New Hugging Face Space 30:18 | Git and GitHub Desktop 31:20 | Using Git from the Terminal 32:02 | Windows Terminal and Ubuntu 32:40 | Working on Your Own Machine 33:45 | Using the Terminal for Copy and Paste 34:00 | Creating a Directory and File 34:26 | Opening the Directory in VS Code 35:06 | VS Code and Creating an app.py File 35:30 | Committing to Hugging Face Spaces 36:03 | Building the Website and Gradio Interface 36:41 | Testing the Interface and Running an App 37:05 | Deep Learning Model in Production 37:17 | Training a Dog or Cat Classifier 37:44 | Kaggle Model and Edit/Reader Views 38:50 | Exporting the Trained Model 39:19 | Colab Model and Downloading the Model 40:07 | Copying the Model to the Hugging Face Spaces Directory 41:09 | Creating a Space and Downloading the Model 41:23 | Making Predictions on a Saved Model 41:37 | Using a Trained Model for Predictions 42:12 | Loading the Learner and Making Predictions 43:23 | Unpickling the Learner and the Predict Method 44:02 | Returning Predictions and Probabilities 44:28 | Creating a Gradio Interface 45:05 | Python Idioms and Gradio Limitations 45:42 | Classifying an Image and Creating the Interface 46:17 | Launching the Interface and Testing 46:52 | The Dogo Story and Model Confidence 48:00 | Training Schedule and Model Confidence 48:18 | Stopping the Interface and Creating a Python Script 48:34 | Copying and Pasting Code into a Script 49:00 | Using Hash Pipe Export and nbdev 49:31 | Exporting the Notebook to a Script 50:00 | Uploading the Script to Hugging Face Spaces 50:44 | Testing the Model in Production 51:09 | Retraining the Model and Model Confidence 51:43 | Community Questions and Epochs 52:15 | Epochs and Error Rate 53:21 | Downloading the Model in Colab 53:59 | Free Resources for GPU Training 54:26 | Installing Python and Jupyter Notebooks 54:42 | Fast Setup and GitHub Repository 55:03 | Cloning the Repository and SetupConda.sh 56:06 | Installing Python and Conda-based Distributions 56:51 | MambaForge and Installing Fast.ai 57:28 | Installing Fast.ai with Mamba 58:11 | Installing nbdev and Using Jupyter 58:43 | Starting Jupyter Notebook 59:17 | Pip and Mamba for Installing Software 59:50 | Asking for Help and Reviewing Steps 1:00:01 | Creating a Space, Interface, and Git Setup 1:00:08 | Mamba and Conda and Dogs vs. Cats 1:00:38 | Pet Breeds and Exported Learner 1:00:47 | Using nbdev and Trying the API 1:00:50 | Free Working Model and Gradio Flexibility 1:01:36 | Streamlit and Building Prototypes 1:02:01 | Building an App and the API 1:02:42 | JavaScript and Front-end Engineering 1:03:10 | JavaScript and Data Scientists 1:03:29 | Understanding JavaScript and the API Endpoint 1:04:13 | Testing the API with Curl 1:04:56 | Creating a Website with TinyPets 1:05:10 | JavaScript App and HTML Code 1:06:01 | Running the App in a Browser 1:06:09 | Basic Steps in JavaScript Code 1:07:12 | Building a Multi-file Version 1:07:27 | Testing the Multi-file Version 1:07:55 | Code for the Multi-file Version 1:08:10 | Community Examples and Get to Know Your Pet 1:08:39 | Combining Models and NLP 1:09:29 | Creating the Website and HTML Files 1:09:41 | Creating a Website and GitHub Pages 1:09:46 | Running the HTML File in a Browser 1:10:38 | JavaScript App and Execution Environment 1:11:02 | Hosting the Website with GitHub Pages 1:11:09 | GitHub Pages and Fast Pages 1:11:44 | Setting Up Fast Pages 1:12:13 | Creating a GitHub Pages Site 1:12:24 | Fastai/Tinypets and GitHub Desktop 1:12:49 | Publishing the Website and Themes 1:13:05 | Config.yaml and Jekyll Themes 1:13:27 | Saving Files and Frontmatter 1:14:09 | Customizing the Website and Forking 1:14:19 | Forking the Repository and Changing the Theme 1:14:48 | Enabling GitHub Pages and Changing the Theme 1:15:13 | Viewing the Updated Website 1:15:30 | JavaScript, Websites, and Web Apps 1:15:57 | Free Resources and Hugging Face Spaces 1:16:07 | Signing Out and Next Lesson 1:16:18 | Natural Language Processing and Model Workings 1:16:29 | Stochastic Gradient Descent and Calculus 1:16:39 | See You Next Time