🔥 Learning Interpretability Tool (LIT)
The Learning Interpretability Tool (
LIT is built to answer questions such as:
- What kind of examples does my model perform poorly on?
- Why did my model make this prediction? Can this prediction be attributed to adversarial behavior, or to undesirable priors in the training set?
- Does my model behave consistently if I change things like textual style, verb tense, or pronoun gender?
LIT supports a variety of debugging workflows through a browser-based UI. Features include:
- Local explanations via salience maps, attention, and rich visualization of model predictions.
- Aggregate analysis including custom metrics, slicing and binning, and visualization of embedding spaces.
- Counterfactual generation via manual edits or generator plug-ins to dynamically create and evaluate new examples.
- Side-by-side mode to compare two or more models, or one model on a pair of examples.
- Highly extensible to new model types, including classification, regression, span labeling, seq2seq, and language modeling. Supports multi-head models and multiple input features out of the box.
- Framework-agnostic and compatible with TensorFlow, PyTorch, and more.
LIT has a website with live demos, tutorials, a setup guide and more.
Stay up to date on LIT by joining the lit-announcements mailing list.
For a broader overview, check out our paper and the user guide.
Documentation
Download and Installation
LIT can be installed via pip or built from source. Building from source is
necessary if you update any of the front-end or core back-end code.
Install from source
Clone the repo and set up a Python environment:
git clone https://github.com/PAIR-code/lit.git ~/lit
# Set up Python environment
cd ~/lit
conda env create -f environment.yml
conda activate lit-nlp
conda install cudnn cupti # optional, for GPU support
conda install -c pytorch pytorch # optional, for PyTorch
# Build the frontend
(cd lit_nlp; yarn && yarn build)Note: if you see an error
running yarn on Ubuntu/Debian, be sure you have the
correct version installed.
pip installation
pip install lit-nlpThe pip installation will install all necessary prerequisite packages for use
of the core LIT package.
It does not install the prerequisites for the provided demos, so you need to install those yourself. See environment.yml for the list of packages required to run the demos.
Running LIT
Explore a collection of hosted demos on the LIT website demos page.
Quick-start: classification and regression
To explore classification and regression models tasks from the popular GLUE benchmark:
python -m lit_nlp.examples.glue_demo --port=5432 --quickstartNavigate to http://localhost:5432 to access the LIT UI.
Your default view will be a small BERT-based model fine-tuned on the Stanford Sentiment Treebank, but you can switch to STS-B or MultiNLI using the toolbar or the gear icon in the upper right.
Quick-start: language modeling
To explore predictions from a pre-trained language model (BERT or GPT-2), run:
python -m lit_nlp.examples.lm_demo --models=bert-base-uncased --port=5432And navigate to http://localhost:5432 for the UI.
Notebook usage
Colab notebooks showing the use of LIT inside of notebooks can be found at google3/third_party/py/lit_nlp/examples/notebooks.
We provide a simple Colab demo. Run all the cells to see LIT on an example classification model in the notebook.
Run LIT in a Docker container
LIT can be run as a containerized app using Docker or
your preferred engine. Use the following shell commands to build the default
Docker image for LIT from the provided Dockerfile, and then run a container
from that image. Comments are provided in-line to help explain each step.
# Build the docker image using the -t argument to name the image. Remember to
# include the trailing . so Docker knows where to look for the Dockerfile
docker build -t lit_app .
# Now you can run LIT as a containerized app using the following command. Note
# that the last parameter to the run command is the value you passed to the -t
# argument in the build command above.
docker run --rm -p 5432:5432 lit-appThe image above defaults to launching the GLUE demo on port 5432, but you can override this using environment variables. See our advanced guide for detailed instructions on using the default LIT Docker image, running LIT as a containerized web app in different scenarios, and how to creating your own LIT images.
More Examples
See lit_nlp/examples. Run similarly to the above:
python -m lit_nlp.examples.<example_name> --port=5432 [optional --args]User Guide
To learn about LIT's features, check out the user guide, or watch this video.
Adding your own models or data
You can easily run LIT with your own model by creating a custom demo.py
launcher, similar to those in lit_nlp/examples. The
basic steps are:
- Write a data loader which follows the
DatasetAPI - Write a model wrapper which follows the
ModelAPI - Pass models, datasets, and any additional components to the LIT server class
For a full walkthrough, see adding models and data.
Extending LIT with new components
LIT is easy to extend with new interpretability components, generators, and more, both on the frontend or the backend. See our documentation to get started.
Pull Request Process
To make code changes to LIT, please work off of the dev branch and
create pull requests
(PRs) against that branch. The main branch is for stable releases, and it is
expected that the dev branch will always be ahead of main.
Draft PRs are encouraged, especially for first-time contributors or contributors working on complex tasks (e.g., Google Summer of Code contributors). Please use these to communicate ideas and implementations with the LIT team, in addition to issues.
Prior to sending your PR or marking a Draft PR as "Ready for Review", please run the Python and TypeScript linters on your code to ensure compliance with Google's Python and TypeScript Style Guides.
# Run Pylint on your code using the following command from the root of this repo
pushd lit_nlp & pylint & popd
# Run ESLint on your code using the following command from the root of this repo
pushd lit_nlp & yarn lint & popdCiting LIT
If you use LIT as part of your work, please cite our EMNLP paper:
@misc{tenney2020language,
title={The Language Interpretability Tool: Extensible, Interactive Visualizations and Analysis for {NLP} Models},
author={Ian Tenney and James Wexler and Jasmijn Bastings and Tolga Bolukbasi and Andy Coenen and Sebastian Gehrmann and Ellen Jiang and Mahima Pushkarna and Carey Radebaugh and Emily Reif and Ann Yuan},
booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
year = "2020",
publisher = "Association for Computational Linguistics",
pages = "107--118",
url = "https://www.aclweb.org/anthology/2020.emnlp-demos.15",
}
Disclaimer
This is not an official Google product.
LIT is a research project and under active development by a small team. There will be some bugs and rough edges, but we're releasing at an early stage because we think it's pretty useful already. We want LIT to be an open platform, not a walled garden, and we would love your suggestions and feedback - drop us a line in the issues.