PyTorch + captum on Colab: Get it to work?

vuhung · September 22, 2025, 4:08am

Hi everyone,

How to get the example below works on Google Colab?

(IMDB_TorchText_Interpret.ipynb)

!pip install torch

!pip install torchtext

!pip install captum

didn’t work.

I am trying to get it worked on Colab.

If I couldn’t, running locally will be the 2nd option.

github.com/pytorch/captum

tutorials/IMDB_TorchText_Interpret.ipynb

master

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Interpreting text models:  IMDB sentiment analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook loads a pretrained CNN model for sentiment analysis on an IMDB dataset. It makes predictions on test samples and interprets those predictions using the Integrated Gradients method.\n",
    "\n",
    "The model was trained using an open source sentiment analysis tutorials described in: https://github.com/bentrevett/pytorch-sentiment-analysis/blob/master/4%20-%20Convolutional%20Sentiment%20Analysis.ipynb with the following changes:\n",
    "\n",
    "- TEXT: set lower=True at initialization and call build_vocab() on the entire training data including validation, to avoid mismatched indices\n",

This file has been truncated. show original

ptrblck · September 22, 2025, 2:00pm

What’s the error you are seeing?

vuhung · September 23, 2025, 1:04am

Thanks for the rep.

You can see the errors by loading the notebook into Colab.

The current code doesn’t work with Colab, of course by looking at pip install for required libraries.