NheLPer
NheLPer is Python package designed to ease behavioral testing of Natural Language Processing models to identify possible capability failures.
1. About the project
Behavioral tests are intended to test a model against some input data while treating as a black box. The aim is to observe the model's reaction against some perturbations that might occur once the model is productionized. For a more detailed explanation on behavioral testing of NLP models I encourage you to read the insightful paper: Beyond Accuracy: Behavioral Testing of NLP models with CheckList
NLPtest provides helper objects for three different aspects:
- easily generate text samples
- test some specific behaviors of your model
- aggregate the tests outcomes of your model
2. Getting started
2.1. Installation
You can directly install NheLPer using pypi:
pip3 install nhelper
2.2. Usage
To help you get the hang of the library we provide three different Notebooks to the user, accessible from
the examples/ folder:
Samples_generation.ipynb: shows you how to easily generate texts using theGeneratorobject.Please_Behave.ipynb: getting familiar with theBehaviorobject.End2End_tests.ipynb: how to run tests and get an overview of your model behavior.
References
Below, you can find resources that were used for the creation of NLPtest as well as relevant resources about behavioral testing.
Metadata
Release files for nhelper 0.1.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| NheLPer-0.1.0.tar.gz | 49.1 kB | Without provenance | Details for NheLPer-0.1.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| NheLPer-0.1.0-py3-none-any.whl | Python 3 | none | any | Details for NheLPer-0.1.0-py3-none-any.whl |
Total release size: 61.8 kB
Release files / NheLPer-0.1.0.tar.gz
| Download URL | NheLPer-0.1.0.tar.gz |
|---|---|
| Size | 49.1 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
92ae854cc78d6d55cc7b4c3d6fa154a844114674f4f8953b43c93d2ddd953bed
|
|
BLAKE2b-256 checksum How to use checksums |
5131ab96d210e5eac0e54aae164310eb70a4fea800ce57bc95e3675e513a1f5c
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
poetry/1.1.13 CPython/3.9.10 Darwin/21.4.0
|
Release files / NheLPer-0.1.0-py3-none-any.whl
| Download URL | NheLPer-0.1.0-py3-none-any.whl |
|---|---|
| Size | 12.7 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
ef6e39c0b9bb574c105ea889dfb1255babe2df6a8be765474e4be17dc4fb1e39
|
|
BLAKE2b-256 checksum How to use checksums |
1d08b3585cf5c8ef07ad7e484465a462d51ce28ec3640fb517a8211bd129a73f
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
poetry/1.1.13 CPython/3.9.10 Darwin/21.4.0
|
Security for nhelper 0.1.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.