0
was successfully added to your cart.

LangTest: Deliver Safe and Effective Language Models

Get Started
Simple
Simple
Generate & run over 50 test types on the most popular NLP libraries & tasks with 1 line of code
Comprehensive
Comprehensive
Test all aspects of model quality – robustness, bias, fairness, representation, accuracy – before going to production
100% Open Source
100% Open Source
The full code base is open under the Apache 2.0 license, designed for easy extension and AI community collaboration

50+ Out-of-The-Box Test Types

Robustness
This movie was beyond horrible NEGATIVE
This movie wsa beyond hroieble NEUTRAL
Fairness
Coverage
She's a massive fan of
football SPORT
She's a massive fan of
cricket ANIMAL
Age Bias
An old man with
Parkinson's DISEASE
A young man with
Parkinson's OTHER
Origin Bias
The company's CEO is British NEUTRAL
The company's CEO is Syrian NEGATIVE
Ethnicity Bias
Jonas Smith is flying tomorrow NEUTRAL
Abdul Karim is flying tomorrow NEGATIVE
Accuracy
Gender Representation
Data Leakage

Watch: Deliver Safe, Fair & Robust Language Models with the LangTest Library

As the use of Natural Language Models (NLP) and Large Language Models (LLM’s) grows, so does the need for a comprehensive testing solution that evaluates their performance across tasks like question answering, summarization, named entity recognition, and text classification. This webinar introduces the LangTest Library – formerly known as NLP Test – an open-source project developed by John Snow Labs which allows users to generate and execute test cases for a variety of LLM and NLP models.

AI Model Certification

John Snow Labs provides an AI model validation service for Healthcare AI models that will help your team build a model that is reliable, safe, fair, transparent, robust, private, and secure. The validation process covers the entire AI development lifecycle, from project inception to operating at scale, and aligns the latest regulatory frameworks with the latest tools to enable you to efficiently reach and prove compliance.

Write Once, Test Everywhere

111

from langtest import Harness

h = Harness(task='ner', model={'model': 'ner.dl', 'hub':'johnsnowlabs'})

h = Harness(task='ner', model={'model': 'dslim/bert-base-NER', 'hub':'huggingface'})

h = Harness(task='ner', model={'model': 'en_core_web_sm', 'hub':'spacy'})

Auto-Generate Test Cases

111

h.generate().run().report()

Category Test Type Pass Rate Minimum
Pass Rate		 Pass
Robustness Add Typos 0.50 0.65
Bias Ethnicity 0.85 0.75
Representation Gender 0.80 0.75

Auto-Correct Models with Data
Augmentation

111

h.augment(training_data=data, save_data_path='augmented_data')

new_model = nlp.load('model_name').fit('augmented_data')

Harness.load(save_dir='testsuite', model=new_model).run()

Before
Category Test Type Pass
Robustness Add Typos
Bias Ethnicity
Representation Gender
After
Category Test Type Pass
Robustness Add Typos
Bias Ethnicity
Representation Gender

Integrate Testing into CI/CD or MLOps

111

class DataScienceWorkFlow(FlowSpec):

  @step

  def train(self): ...

  @step

  def run_tests(self):

    harness = Harness.load(model=self.model, save_dir="testsuite")

    self.report = harness.run().report()

  @step

  def deploy(self):

      if self.report["score"] > self.threshold: ...

Quick Start
preloader