Issues

• Datasets/papers to explore

  mrdbourke/nutrify#74 · Filter by author mrdbourke opened on Mar 1, 2023

• Dashboard creation for data exploration

  mrdbourke/nutrify#72 · Filter by author mrdbourke opened on Feb 13, 2023
  7

• Change autocorrect_and_merge_labels pipeline to use OpenCLIP rather than generic CLIP

  mrdbourke/nutrify#71 · Filter by author mrdbourke opened on Feb 12, 2023

• Autolabelling object detection and semantic segmentation

  mrdbourke/nutrify#70 · Filter by author mrdbourke opened on Feb 9, 2023

• Go through all_images on Google Storage and resize them to take up less space

  mrdbourke/nutrify#69 · Filter by author mrdbourke opened on Feb 8, 2023
  1

• Update main README.md

  mrdbourke/nutrify#66 · Filter by author mrdbourke opened on Jan 31, 2023

• Create a way to generate images to boost inaccurate classes

  database
  mrdbourke/nutrify#65 · Filter by author mrdbourke opened on Jan 31, 2023
  1

• Autolabelling: Setup a data collection pipeline (e.g. what happens when new data comes in?)

  mrdbourke/nutrify#64 · Filter by author mrdbourke opened on Jan 31, 2023
  4

• Simplify training scripts (e.g. merge engine.py -> train.py)

  mrdbourke/nutrify#63 · Filter by author mrdbourke opened on Jan 31, 2023

• Create export.py (or something similar) for exporting models from PyTorch -> CoreML

  mrdbourke/nutrify#62 · Filter by author mrdbourke opened on Jan 31, 2023
  1

• Clean scripts and add docstrings to code in foodvision

  mrdbourke/nutrify#61 · Filter by author mrdbourke opened on Jan 31, 2023

• Start breaking down class_names into more easily understandable foods

  mrdbourke/nutrify#59 · Filter by author mrdbourke opened on Jan 30, 2023
  2

• Add image_source tags to all images in the database

  databaselabelling
  mrdbourke/nutrify#58 · Filter by author mrdbourke opened on Jan 30, 2023

• Profile the training/testing code to see where the bottlenecks are

  training
  mrdbourke/nutrify#56 · Filter by author mrdbourke opened on Jan 25, 2023

• Only validate once per X epochs

  training
  mrdbourke/nutrify#55 · Filter by author mrdbourke opened on Jan 25, 2023

• Try using polars for fasting data loading or even use Hugging Face Datasets

  training
  mrdbourke/nutrify#54 · Filter by author mrdbourke opened on Jan 25, 2023

• Clean/remove duplicate images with fastdup

  mrdbourke/nutrify#53 · Filter by author mrdbourke opened on Jan 23, 2023
  3

• Use Cleanlab to clean labels in dataset

  labelling
  mrdbourke/nutrify#52 · Filter by author mrdbourke opened on Jan 17, 2023

• Create a database (PostgreSQL?) for labels

  databaselabelling
  mrdbourke/nutrify#51 · Filter by author mrdbourke opened on Jan 12, 2023
  1

• Create a script to import the "100 most wrong predictions" and then export them as labelling task in Label Studio

  labellingmachine learning
  mrdbourke/nutrify#47 · Filter by author mrdbourke opened on Dec 14, 2022
  2

• Create a function/script to create different models for different purposes (e.g. big, medium, small)

  machine learning
  mrdbourke/nutrify#45 · Filter by author mrdbourke opened on Dec 14, 2022

• Add a way to save to storage the images uploaded to Nutrify

  mrdbourke/nutrify#43 · Filter by author mrdbourke opened on May 11, 2022
  3

• Make the loading images function faster (this should appear on page load)

  mrdbourke/nutrify#41 · Filter by author mrdbourke opened on Mar 23, 2022
  2

• Database update: make it so a new food gets added via FDC ID -> information pulled into database

  mrdbourke/nutrify#40 · Filter by author mrdbourke opened on Mar 22, 2022
  2

• Model/class names not lined up + some classes are missing FDC data ("Egg Tart", "Fries", "Hamimelon")

  mrdbourke/nutrify#39 · Filter by author mrdbourke opened on Mar 18, 2022
  3