Skip to content
This repository has been archived by the owner on Jun 1, 2019. It is now read-only.

Latest commit

 

History

History
70 lines (58 loc) · 3.59 KB

CHANGELOG.md

File metadata and controls

70 lines (58 loc) · 3.59 KB
Raw

Changelog

  • Better support for displaying bounding boxes in object detection

  • René updates

    • final UI fixes

  • Shane updates

    • model metrics available as a separate component, nav'd to by clicking the metrics link
    • browser back button takes you back to playground
    • nested table for epoch label information, where available
    • color-coded confusion matrix cells
    • for image/multi-image clicking on a non-zero confusion matrix cell opens the list of examples for that cell
    • epochs display for multi-image are split across labels, with child rows for positive/negative for each label
    • additional info for object detection (precision, recall) logged to browser console where displaying on screen won't make sense.

  • René updates

    • Got rid of modelmetrics and learning curve classes, replaced with String
    • Added alternative support for einstein cert as File to make transition easier

  • Shane updates

    • move to certificate based authentication
    • label/model count in badges on playground
    • model stats (confusion, f1, etc)
    • usage/limits component/tab
    • dataset cards can grow vertically with label count
    • separate playground for multi-image detection
    • refactoring/consolidation of components
    • build language models from Salesforce data
    • example processBuilder implementation of language predictions + feedback (inactive, don't worry!)
    • predictions from image urls
    • intelligent, high-contrast color coding for object detection
    • scalable (responsive) object detection image display
    • metrics on models in a modal
    • failure messages for model training
    • model retrain of most recent model (keep same modelId)
    • friendly names (model names) for model selection dropdown

  • 2.5.0

    • Bubbling up errors to Lighting components in the Playground Apex controller
    • Adjusting bounding box display for object detection
    • Fixed Null Pointer Exception when user opens tab without data
    • Addes .yaml configuration for deploy-to-sfdx button

  • 2.4.5 - 11/25/17

    • Cleaned up permissions in Admin profile
    • Fixed display of setup page

  • 2.4.0 - 11/16/17

    • Added example Process Builder implementation for image classification
    • Increased HTTP timeout to 30 seconds
    • Raised image upload limit to 5MB

  • 2.3.0 - 10/16/17

    • Adding Einstein Image Detection (beta)
    • Minor UI changes

  • 2.2.0 - 09/13/17

    • Added setup UI/information to playground - Link
    • Restrict file size upload for image classification - Link
    • Run image prediction after image has been selected - Link
    • UI refinements - Link Link

  • 2.1.0 - 09/08/17

    • Added support for Salesforce orgs with namespace - Link
    • Updated UI example with NEW PLAYGROUND - Link
    • Moved training parameters from String to Apex Object - Link

  • 2.0.0 - 08/22/17

    • Initial version