--- title: "Unitxt Metric" emoji: 📈 colorFrom: pink colorTo: purple sdk: static app_file: README.md pinned: false ---
Image Description
[![Button](https://img.shields.io/badge/Video-pink?style=for-the-badge)](https://unitxt.readthedocs.io/en/latest/_static/video.mov) [![Button](https://img.shields.io/badge/Documentation-pink?style=for-the-badge)](https://unitxt.readthedocs.io/en/latest/docs/introduction.html) [![Button](https://img.shields.io/badge/Demo-pink?style=for-the-badge)](https://unitxt.readthedocs.io/en/latest/docs/demo.html) [![Button](https://img.shields.io/badge/Tutorial-pink?style=for-the-badge)](https://unitxt.readthedocs.io/en/latest/docs/adding_dataset.html) [![Button](https://img.shields.io/badge/Paper-pink?style=for-the-badge)](https://arxiv.org/abs/2401.14019) [![Button](https://img.shields.io/badge/Catalog-pink?style=for-the-badge)](https://unitxt.readthedocs.io/en/latest/catalog/catalog.__dir__.html) [![Button](https://img.shields.io/badge/Contributors-pink?style=for-the-badge)](https://github.com/IBM/unitxt/blob/main/CONTRIBUTING.md) [![Button](https://img.shields.io/badge/PyPi-pink?style=for-the-badge)](https://pypi.org/project/unitxt/) In the dynamic landscape of generative NLP, traditional text processing pipelines limit research flexibility and reproducibility, as they are tailored to specific dataset, task, and model combinations. The escalating complexity, involving system prompts, model-specific formats, instructions, and more, calls for a shift to a structured, modular, and customizable solution. Addressing this need, we present Unitxt, an innovative library for customizable textual data preparation and evaluation tailored to generative language models. Unitxt natively integrates with common libraries like HuggingFace and LM-eval-harness and deconstructs processing flows into modular components, enabling easy customization and sharing between practitioners. These components encompass model-specific formats, task prompts, and many other comprehensive dataset processing definitions. The Unitxt-Catalog centralizes these components, fostering collaboration and exploration in modern textual data workflows. Beyond being a tool, Unitxt is a community-driven platform, empowering users to build, share, and advance their pipelines collaboratively. # [![version](https://img.shields.io/pypi/v/unitxt)](https://pypi.org/project/unitxt/) ![license](https://img.shields.io/github/license/ibm/unitxt) ![python](https://img.shields.io/badge/python-3.8%20|%203.9-blue) ![tests](https://img.shields.io/github/actions/workflow/status/ibm/unitxt/library_tests.yml?branch=main&label=tests) [![codecov](https://codecov.io/gh/IBM/unitxt/branch/main/graph/badge.svg?token=mlrWq9cwz3)](https://codecov.io/gh/IBM/unitxt) ![Read the Docs](https://img.shields.io/readthedocs/unitxt) [![downloads](https://static.pepy.tech/personalized-badge/unitxt?period=total&units=international_system&left_color=grey&right_color=green&left_text=downloads)](https://pepy.tech/project/unitxt) # https://github.com/IBM/unitxt/assets/23455264/baef9131-39d4-4164-90b2-05da52919fdf ### 🦄 Currently on Unitxt Catalog ![NLP Tasks](https://img.shields.io/badge/NLP_tasks-48-blue) ![Dataset Cards](https://img.shields.io/badge/Dataset_Cards-537-blue) ![Templates](https://img.shields.io/badge/Templates-265-blue) ![Formats](https://img.shields.io/badge/Formats-23-blue) ![Metrics](https://img.shields.io/badge/Metrics-136-blue) ### 🦄 Run Unitxt Exploration Dashboard To launch unitxt graphical user interface first install unitxt with ui requirements: ``` pip install unitxt[ui] ``` Then launch the ui by running: ``` unitxt-explore ``` # 🦄 Example This is a simple example of running end-to-end evaluation in self contained python code over user data. See more examples in examples subdirectory. ```python from unitxt import get_logger from unitxt.api import evaluate, load_dataset from unitxt.blocks import Task, TaskCard from unitxt.inference import HFPipelineBasedInferenceEngine from unitxt.loaders import LoadFromDictionary from unitxt.templates import InputOutputTemplate, TemplatesDict from unitxt.text_utils import print_dict logger = get_logger() # Set up question answer pairs in a dictionary data = { "test": [ {"question": "What is the capital of Texas?", "answer": "Austin"}, {"question": "What is the color of the sky?", "answer": "Blue"}, ] } card = TaskCard( # Load the data from the dictionary. Data can be also loaded from HF, CSV files, COS and other sources using different loaders. loader=LoadFromDictionary(data=data), # Define the QA task input and output and metrics. task=Task( input_fields={"question": str}, reference_fields={"answer": str}, prediction_type=str, metrics=["metrics.accuracy"], ), ) # Create a simple template that formats the input. # Add lowercase normalization as a post processor on the model prediction. template = InputOutputTemplate( instruction="Answer the following question.", input_format="{question}", output_format="{answer}", postprocessors=["processors.lower_case"], ) # Verbalize the dataset using the template dataset = load_dataset(card=card, template=template) test_dataset = dataset["test"] # Infer using flan t5 base using HF API # can be replaced with any prediction code, # including the built in WMLInferenceEngine and OpenAiInferenceEngine. model_name = "google/flan-t5-base" inference_model = HFPipelineBasedInferenceEngine( model_name=model_name, max_new_tokens=32 ) predictions = inference_model.infer(test_dataset) evaluated_dataset = evaluate(predictions=predictions, data=test_dataset) # Print results for instance in evaluated_dataset: print_dict( instance, keys_to_print=[ "source", # input to the model "prediction", # model prediction "processed_prediction", # model prediction after post processing "references", # reference answer "score", # scores (per instance and global) ], ) ``` # 🦄 Contributors Please install Unitxt from source by: ```bash git clone git@github.com:IBM/unitxt.git cd unitxt pip install -e ".[dev]" pre-commit install ``` # 🦄 Citation If you use Unitxt in your research, please cite our paper: ```bib @inproceedings{bandel-etal-2024-unitxt, title = "Unitxt: Flexible, Shareable and Reusable Data Preparation and Evaluation for Generative {AI}", author = "Bandel, Elron and Perlitz, Yotam and Venezian, Elad and Friedman, Roni and Arviv, Ofir and Orbach, Matan and Don-Yehiya, Shachar and Sheinwald, Dafna and Gera, Ariel and Choshen, Leshem and Shmueli-Scheuer, Michal and Katz, Yoav", editor = "Chang, Kai-Wei and Lee, Annie and Rajani, Nazneen", booktitle = "Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 3: System Demonstrations)", month = jun, year = "2024", address = "Mexico City, Mexico", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2024.naacl-demo.21", pages = "207--215", abstract = "In the dynamic landscape of generative NLP, traditional text processing pipelines limit research flexibility and reproducibility, as they are tailored to specific dataset, task, and model combinations. The escalating complexity, involving system prompts, model-specific formats, instructions, and more, calls for a shift to a structured, modular, and customizable solution.Addressing this need, we present Unitxt, an innovative library for customizable textual data preparation and evaluation tailored to generative language models. Unitxt natively integrates with common libraries like HuggingFace and LM-eval-harness and deconstructs processing flows into modular components, enabling easy customization and sharing between practitioners. These components encompass model-specific formats, task prompts, and many other comprehensive dataset processing definitions. The Unitxt Catalog centralizes these components, fostering collaboration and exploration in modern textual data workflows. Beyond being a tool, Unitxt is a community-driven platform, empowering users to build, share, and advance their pipelines collaboratively. Join the Unitxt community at https://github.com/IBM/unitxt", } ``` Unitxt emoji designed by [OpenMoji](https://openmoji.org/#) - the open-source emoji and icon project. License: [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/#)