How can I make my Bear Classifier into a website, if I am using Google Colab

himanshu_j · April 5, 2021, 3:23pm

How can I make my Bear Classifier into a website, if I am using Google Colab.
I didn’t find any suitable way using ngrok or voila. Please Help!

sebastian12 · April 15, 2021, 12:49am

Hello, I have kind of a similar question . I have used Dash before to make an interactive website using python, I thought I could use Dash and fastai together to make a web app but so far I haven´t been able to do it. I wonder if theres is some other way to upload images , because I haven´t been able to use:

btn_upload = widgets.FileUpload()
out_pl = widgets.Output()
lbl_pred = widgets.Label()

widgets are for jupyter notebooks or google colab but not for pycharm.

I´ve tried using:
btn_upload = ‘upload-image’
out_pl = ‘output-image-upload’
lbl_pred = widgets.Label() <-- I really don´t know what to change this for

but I haven´t been able to have success.

Pugh201 · May 5, 2021, 8:12pm

It seems to involve saving pretrained weights, freezing the network and adding the new classes and then continuing the training. Similar to the transfer learning we do when we use the Resnet model.

I haven’t done it personally but can see a need in the future for one of my solutions.

zerotosingularity · May 6, 2021, 8:01am

Have you followed the steps in the Fastbook Production Notebook?

github.com

fastai/fastbook/blob/master/02_production.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#hide\n",
    "! [ -e /content ] && pip install -Uqq fastbook\n",
    "import fastbook\n",
    "fastbook.setup_book()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [

This file has been truncated. show original

There are a couple of threads that might put you in the right direction:

https://forums.fast.ai/search?q=voila

Alternatively, you can deploy your model to the web, iOS and Android using the SeeMe.ai quick guide:
https://course.fast.ai/deployment_seeme_ai
No setup required, and models are converted to also run locally/offline on the mobile devices.
Disclaimer, I’m the creator or SeeMe.ai, so I’m obviously biased, but here to help.