Scaling Synthetic Data Creation with 1 Billion Personas | PersonaHub Dataset Explained Argilla 2024-07-10 | Welcome to another episode of Data Explorer by Argilla! 🎥🚀 In this episode, we’re diving into the Persona Hub dataset, introduced in the paper “Scaling Synthetic Data Creation with 1 Billion Personas” by Xin Chan et al from the Tencent AI Lab.This dataset focuses on increasing the variety in synthetic datasets by using personas. By assigning a persona to a large language model (LLM), we can create more diverse and realistic responses to instructions. The paper proposes a method to create these personas from world knowledge and public texts from the web.Resources:- Dataset repo: huggingface.co/datasets/proj-persona/PersonaHub- Notebook to upload to Argilla: colab.research.google.com/drive/1axpG_6s0DE6X-k9JPzkCXQwAeEbahG1Z?usp=sharing- Paper: huggingface.co/papers/2406.20094- Argilla Instance: huggingface.co/spaces/argilla/data-explorers
Synthetic Data Generator - Build Datasets Using Natural Language Argilla 2024-12-16 | Synthetic data can have enormous benefits by training a smaller, more efficient model, instead of relying on an LLM.Hugging Face now offers a free no-code UI to start with synthetic data generation. Blog: huggingface.co/blog/synthetic-data-generatorPlayground: huggingface.co/spaces/argilla/synthetic-data-generatorGitHub: github.com/argilla-io/synthetic-data-generator
FineWeb2 collaborative sprint: how to annotate Argilla 2024-12-13 | In this video you'll learn how you can go about annotating some records in the FineWeb2 collaborative annotation sprint launched by Hugging Face and Argilla. You will see some examples and we'll guide you on how to decide which label fits best. Do you want to give it a try? - Start annotating here: huggingface.co/spaces/data-is-better-together/fineweb-c- Join our chat: huggingface.co/spaces/HuggingFaceFW/discussion
Start a token classification project on the Hugging Face Hub with Argilla, GliNER and NuExtract LLM Argilla 2024-11-18 | Show a standard workflow for a token classification including zero-shot suggestions and model fine-tuning with Argilla, GliNER and Spanmarker.- Deploy argilla on Spaces- Vibe check your dataset- Configure and create an Argilla dataset- Add records- Add zero-shot suggestions with GliNER- Evaluate model suggestions in Argilla- Train a SpanMarker modelHope to see all of you guys there and looking forward to your questions and AI use cases.You can find an overview of the shared documents here: drive.google.com/drive/folders/1JECQm1SotvlD04L-oXJFQq6ttxm3lv0c?usp=sharingSignup for coming meetups here: https://lu.ma/d720wy9f
Start a text classification project on the Hugging Face Hub with Argilla and SetFit Argilla 2024-10-15 | Learn about a standard workflow for text classification with model fine-tuning in Argilla using SetFit.- Deploy argilla on Spaces- Vibe check your dataset- Configure and create an Argilla dataset- Add records- Add zero-shot suggestions- Evaluate model suggestions in Argilla- Train a SetFit modelYou can find an overview of the shared documents here: drive.google.com/drive/folders/1JECQm1SotvlD04L-oXJFQq6ttxm3lv0c?usp=sharingSignup for coming meetups here: https://lu.ma/d720wy9f
Image projects on Hugging Face: from fine-tuning CLIP models to synthetic image datasets Argilla 2024-09-16 | We discussed not one, not two but 3 notebooks on image stuff.1. finetune an CLIP retrieval model with sentence transformers2. use ColPali+ Qwen VL for RAG and log the results to Argilla3. image-generation preference: creating multi-modal preference datasets for free using Hugging Face inference endpoints.You can find an overview of the shared documents here: drive.google.com/drive/folders/1JECQm1SotvlD04L-oXJFQq6ttxm3lv0c?usp=sharingSignup for coming meetups here: https://lu.ma/d720wy9f
What is distilabel? A brief feature overview. Argilla 2024-08-30 | We discussed distilabel during one of our meetups and also gave a brief feature overview.youtu.be/quU0DGml5Vo
Generating and cleaning a preference dataset for DPO / ORPO with LLMs and distilabel Argilla 2024-08-30 | We discussed:The essentials of building a distilabel pipeline by exploring two key use cases: cleaning an existing dataset and generating a preference dataset for DPO/ORPO. You’ll also learn how to make the most of it, integrating Argilla to gather human feedback and improve its quality.This session is perfect for youif you’re getting started with distilabel or synthetic dataif you want to discover new functionalitiesif you want to provide us with new feedbackYou can find an overview of the shared documents here: drive.google.com/drive/folders/1JECQm1SotvlD04L-oXJFQq6ttxm3lv0c?usp=sharingSignup for coming meetups here: https://lu.ma/d720wy9f
Optimizing RAG Pipelines by fine-tuning custom embedding models on synthetic data with ZenML Argilla 2024-08-09 | ZenML is an MLOps + LLMOps framework that makes your infrastructure and workflow metadata accessible to data science teams. Orchestrate your machine learning and training pipelines with ZenML in combination with powerful tools like distilabel and Argilla. Alex Strick van Linschoten will showcase how you can use synthetic data generated by distilabel to bootstrap embedding model fine-tuning and then use human feedback in Argilla to iteratively and continuously improve the model performance.You can find an overview of the shared documents here: drive.google.com/drive/folders/1JECQm1SotvlD04L-oXJFQq6ttxm3lv0c?usp=sharingSignup for coming meetups here: https://lu.ma/d720wy9f
ZenML a way to streamline your complex projects with ease Argilla 2024-08-09 | it's a great way to keep track of all of the things that you're doing in what can, Often become like quite quite quite complex.Especially when you have lots of different teams that are involved. Potentially if you deployed your model, we're not doing it here. But the deployment and you can keep track of that particular deployment of a model that would live here. And then you have metadata relating to your particular pipelines.You have eval results which are displayed here. The dimensions of these Matryoshka loss functions are stored here. Some stuff about my local computer that I was running this on.
cosine similarity as proxy for quality of sentence pair data Argilla 2024-08-09 | The positive query has a higher similarity, which is better because then it would have been found by the model and comparatively, the negative query should have a relatively lower similarity because then it wouldn't have been found or isn't competing with the positive that much. Then what I will do is based on the assumption that you actually want to either maximize or minimize this retrieval chance, so to say, you can actually. by just looking at your data and making some basic assumptions, you already filter out some of these outliers.
optimizing RAG by choosing the right model Argilla 2024-08-09 | So let's go for a small model, but not too small. And then what we can do is for example, focus on the task of retrieval because we're working with a rack pipeline.You can also do re ranking and then you end up with apparently these cool snowflake open source models. That you can use for free and use out of the box as well. You can still check what the underlying data sets are that they used. For example, if they used it specifically for a medical domain or these kinds of things, it will not show up because it's like the average score that we're looking at.Which you might be able to look at some other scores within the official dashboard as well.
model pooling for diverse synthetic data generation Argilla 2024-08-09 | You can also play around with using a variety of different LLMs in the generation because you can get sameness in the response or sameness in the kinds of ways that the responses happen and going from The not so great LLMs to like, more capable ones and having a mix.I'm not sure whether this label allows you to combine multiple LLMs within the same generation pipeline. Yeah, you can. So there's this model pooling concept where I guess it's based on also the ultra feedback paper where they just had a group of a pool of different models.I'm not sure whether this label allows you to combine multiple LLMs within the same generation pipeline. Yeah, you can. So there's this model pooling concept where I guess it's based on also the ultra feedback paper where they just had a group of a pool of different models
Ellamind on synthetic data generation with distilabel for pipelining and LLM finetuning Argilla 2024-07-22 | We welcome Daniel from Ellamind and Disco Research to explore their use of distilabel for generating synthetic data. Discover the journey of Ellamind in developing an email reply application for a customer service use case. Daniel dives into the company's origins, the crucial role of synthetic data, and how distilabel pipelines are utilized for fine-tuning models like Mixtral. He highlights their approach to augmenting limited datasets, fine-tuning embedding models, and improving email responses with LLMs. Additionally, Daniel shares insights on self-hosting models versus using APIs, and the benefits of open-source tools, and discusses their contributions to the multilingual MTAP benchmark. Learn how Ellamind is pushing the boundaries of customer service applications through innovative synthetic data generation.00:00 Introduction to the Session01:07 Overview of EleMind and Disco Research02:04 Demo of the Email Reply Application03:10 Need for Synthetic Data04:10 Generating Synthetic Emails07:28 Using Distillable Pipelines11:36 Open Source Contributions and Benchmarks14:33 Q&A Session29:45 Closing RemarksYou can find an overview of the shared documents here: drive.google.com/drive/folders/1JECQm1SotvlD04L-oXJFQq6ttxm3lv0c?usp=sharingSignup for coming meetups here: https://lu.ma/d720wy9f
Javier Alonso on lead optimisation at Idealista Argilla 2024-06-26 | Optimizing Lead Classification at Idealista Using Argilla and GPT-4In this meetup, join David and Javi from Idealista to explore their recent project on optimizing lead classification. Leveraging Argilla and GPT-4, they tackle the issue of filtering thousands of user messages to ensure only relevant and appropriate communications reach real estate agents. The session covers the entire journey from initial problem identification, data extraction, auto labeling using GPT-4, to human verification and model training. Gain insights into the importance of accurate data, innovative labeling strategies, and the technical tools that enabled the success of this project. Tune in to learn how Idealista efficiently manages lead categorization and enhances its operational workflow.00:00 Introduction and Guest Welcome01:40 Overview of Idealista03:47 Understanding the Problem: Lead Classification10:31 Data Collection and Preparation17:00 Auto Labeling with Large Language Models24:42 Introducing Auto Labels with Large Language Models24:51 Exploring Argilla's Suggestions Feature26:28 Batch Labeling for Efficient Processing27:41 Iterative Record Increase and Model Training30:11 Fine-Tuning Transformers for Text Classification33:21 Conclusions and Key Takeaways38:34 Q&A Session: Language Models and Project Insights48:48 Final Remarks and Future Plans
Exploring the PRISM Dataset: Conversations, Insights, and Model Performance Argilla 2024-06-26 | 📍📍📍 Welcome to the first Data Explorer video by Argilla! In this series, we’ll dive into fascinating datasets available on the hub, uncovering their origins, purposes, and potential applications.In this inaugural episode, we explore the PRISM dataset created by Hannah Rose, Kirk, and colleagues. This dataset is designed to offer diverse forms of alignment based on people from different demographics, collected through two stages: surveys and guided conversations with large language models (LLMs).Check out the detailed model card on the hub for more information and start exploring the PRISM dataset today!🔗 Visit the hub to explore the PRISM dataset and find out more: huggingface.co/datasets/HannahRoseKirk/prism-alignment📓 If you want to try this out, I shared the notebook here:colab.research.google.com/drive/1JXzwkfmxPjShhRjN0UvC4-1ZEL-ZNXdE?usp=sharing
Ben Burtenshaw on the Argilla 2.0 SDK refactor Argilla 2024-06-16 | Join us for a presentation on the newly refactored Argilla SDK. This session will cover:Why the Change: Over the past year, Argilla shifted from task-centric datasets to extensible ones, targeting complex tasks. Feedback from our community highlighted the need for simpler, more flexible tools.How We Improved: The new SDK focuses purely on extensible datasets, offering a streamlined, user-friendly experience. Key changes include a new client object for server interactions, simplified dataset management, and enhanced task definitions.Discover how these updates can help you create high-quality datasets more efficiently. The session will focus on these core workflows:- Simplified server interactions with the new client object- Efficient management of extensible datasets- Task definition using the new Settings class- Methods for adding, updating, and querying recordsYou can find an overview of the shared documents here: drive.google.com/drive/folders/1JECQm1SotvlD04L-oXJFQq6ttxm3lv0c?usp=sharingSignup for coming meetups here: https://lu.ma/d720wy9f
Louis Guitton on NER with Argilla Argilla 2024-05-29 | In this community meetup we had Louis Guitton, an AI Solutions Architect and a long community member that explained the basics on how doing NER with Argilla. He explained the features introduced in the latest releases for doing token and span classification, and listed resources such as tools and models to get quickly started with NER.You can find an overview of the shared documents, chat and QnA here: drive.google.com/drive/folders/1X0-92flvPGi1K_7t3HMJ7lultPW5mcwD?usp=drive_linkSignup for coming meetups here: https://lu.ma/d720wy9f
Weights & Biases on Wandbot Argilla 2024-05-29 | In this community meetup we had Ayush and Bharat from Weights & Biases talking about Wandbot, a bot they develop to internally help users using Weights & Biases platform. They explained us how they went from their initial prototype born in an internal hackathon to the version deployed in production today, and how they used Argilla for the evaluation.You can find an overview of the shared documents, chat and QnA here: drive.google.com/drive/folders/1AxBfSsGZItU5VCvyuoF6qR1axekbttaI?usp=drive_linkSignup for coming meetups here: https://lu.ma/d720wy9f
Datamaran on using Argilla in MLOps workflows for ESG governance Argilla 2024-05-08 | In this community meetup we had Vincent and Martin from Datamaran, a company Argilla has been collaborating with for a few years. They explained us how they use Argilla for annotating text data used to train their machine learning models, and how they are managing the lifecycle of these models applied to the context of ESG.You can find an overview of the shared documents, chat and QnA here: drive.google.com/drive/folders/1JOov70Q0Fj2nloH-A6ZeiKa7PHHGn7Ci?usp=drive_linkSignup for coming meetups here: https://lu.ma/d720wy9f
Understanding and reproducing DEITA with MantisNLP using distilabel=1.0.0 Argilla 2024-04-02 | We are very eager to announce our collaborative community meetup with MantisNLP. We've been working on this wonderful blog series on RLHF and alternatives but in this community edition we will discuss the following:- Dissecting the Deita paper is fundamental to getting high-quality data through AI feedback.- Using distilabel==1.0.0 for a faithful reproduction of the Deita paper.MantisNLP has been a close friend of Argilla so we are really looking forward to this great event. Don't worry, we are also working on an awesome blog about the ORPO paper for our blog series.You can find an overview of the shared documents, chat and QnA here: drive.google.com/drive/folders/1FqEXOTQnz3bvr8Lz7MKt3K8Snsot0G9q?usp=drive_linkSignup for coming meetups here: https://lu.ma/d720wy9f
Elad Levi on AutoPrompt and intent-based prompt calibration and prompt engineering Argilla 2024-03-17 | Large language models (LLMs) are highly effective for many natural language processing tasks when given the right prompts. However, finding an optimal prompt is challenging due to LLMs' notable sensitivity to the prompt input. Additionally, the frequent updates of proprietary models and the emergence of unforeseen edge cases after deployment often requires continuous manual prompt refinement.In this talk, we explore recent methods in prompt optimization and their challenges. We introduce a novel method for automatic prompt engineering that interactively refines prompts based on user intent through synthetic data generation, and mitigates the scarcity of high-quality benchmarks for real-world applications. We will also present our new open-source system for prompt optimization, featuring powerful capabilities such as prompt distillation, prompt squashing, and synthetic benchmark creation.AutoPrompt: github.com/Eladlev/AutoPromptYou can find an overview of the shared documents, chat and QnA here: drive.google.com/drive/folders/1Lrf6dSnNQPkHWHPjU51MOqRCntW3B-gt?usp=drive_linkSignup for coming meetups here: https://lu.ma/d720wy9f
Daniel van Strien on the Hugging Face hub and synthetic creation of a DPO dataset for Haiku Argilla 2024-03-02 | We are really excited about the next community meetup with Daniel van Strien from Hugging Face. For the people who don't know him, Daniel van Strien, is Hugging Face's machine learning librarian and he has also been one of the leaders behind the push for our Hugging Face prompt-collective :)We discussed:- The work Hugging Face is doing to make the Hub better for data.- His beautiful haiku-dpo project which aims to make LLMs better at writing Haiku. The long-term goal is to see if DPO can be used to help make LLMs match an individual's aesthetic preferences. The project was created using distilabel and Argilla.- Why data is better together and we will share some thoughts and preliminary results from our prompt collective efforts.Signup for coming meetups here: https://lu.ma/d720wy9f
Seth Levine on the usage of SetFit and BerTopic for unsupervised clustering Argilla 2024-02-15 | We are happy to invite Seth Levine from the well-known learning from machine learning podcast to the stage to present his implementation of Argilla at Loris.ai.He will walk us through:- What is Loris- What NLP/ML challenges do they face- How they use unsupervised learning to help them with text classification- How they use Argilla to help them quickly validate and gain labelled samples and get classifiers out quicker- Evaluation- Future use-casesAdditionally, we will talk about the progress of the Hugging Face community annotation effort called "prompt collective" and explain why data is much better together. huggingface.co/spaces/DIBT/prompt-collectiveYou can find an overview of the shared documents, chat and QnA here: drive.google.com/drive/folders/15-Eh4RRRLu8Wwd_HOp2yt1f0ANJkR0pS?usp=drive_linkSignup for coming meetups here: https://lu.ma/d720wy9f
Red Cross 510 on NLP for good with SetFit for chat message classification Argilla 2024-02-05 | Within this awesome edition, we welcomed 510, an initiative from the Red Cross of the Netherlands, which is an NGO that was created to support Red Cross operations with data decisions. We discussed- How the Red Cross 510 has been using Argilla for redirecting requests from refugees of the Ukrainian crisis to streamline the support processes of the Red Cross.- How to clean an existing preference dataset using distilabel.Content: drive.google.com/drive/folders/1F1BrkIaxtV-fzQPDQsKG14hy3hE9Tm2y?usp=drive_linkSignup for coming meetups here: https://lu.ma/d720wy9f
Prolific on workload distribution, LLM preference data annotation and Phi2 fine-tune Colab Argilla 2024-01-19 | Our first official collaboration for community meetup with Prolific.ai! Prolific is a company that provides ethically sourced annotators for research and AI projects across diverse demographics and they just announced an integration with Argilla! We discussed:- Distributing annotation work for LLM preference data- The integration of Prolific.ai with Argilla- Fine-tuning phi2 using DPO, LoRa and quantization on a Google Colab GPU (A100)Content used:drive.google.com/drive/folders/12Y9QPqbotRuaqRJVXllCLPX75ELv7Sj-?usp=drive_linkProlific offer:docs.google.com/document/d/1yffaDs8Lu6KIyQTKExruL8Jcuamk5X815aw-1Heau-g/edit?usp=sharingSignup for coming meetups here: https://lu.ma/d720wy9f
Pitching AI to your boss, SLMs vs LLMs and contributing to open source projects Argilla 2024-01-05 | In the first meetup of 2024, we kept it low-key and held an open conversation with several community members.We discussed:- pitching AI to your boss- SLMs vs LLMs- hosting LLMs for inference- contributingSignup for coming meetups here: https://lu.ma/d720wy9f
Kickstart NLP with synthetic data and running LLMs on Google Colab using vLLM Argilla 2023-12-28 | We have been pushing releases within distilabel for synthetic preference data generation. This is why show this Argilla NLP meetup how we can use a free Google Colab GPU to create synthetic data for kickstarting We discussed:- Deploy and LLM using vLLM of Google Colab- How to generate synthetic data- Risks and biases in synthetic dataYou can find an overview of the Google Colab and slides here: - drive.google.com/drive/u/0/folders/1sWRgG2TrYcIikjUWaVVzS7OtaVLyOv2eSignup for coming meetups here: https://lu.ma/d720wy9f
How we cleaned OpenBMB UltraFeedback and Notus Argilla 2023-12-07 | The Notus launch was great and discussing it during our community meetup was great too. Also a special thanks goes out to our first guest speaker Jonny Tran from the Bill and Melinda gates foundation.We discussed:- How we curated the UltraFeedback Synthetic LLM dataset- How we trained Notus 7B OSS LLM- How to use Notus in Hugging Face (inference endpoints + transformers) and distilabel- A guest presentation about a table editor and pdf viewer fork of Argilla You can find an overview of the shared documents, chat and QnA here: app.session.com/history/W8uB9RHzPfgEKk2aqizzywK5DrF_2de6qntINotus blog: argilla.io/blog/notus7bColab data curation: colab.research.google.com/drive/1i20b1ul1MdEdFrJCL9TnsD97uQ71T4_x#scrollTo=aqy4LXeds-7DColab notus usage: colab.research.google.com/drive/1op2rVz1dn_4UJYHsqNlGYhZXOrrZjSyP?usp=sharingFork: github.com/JonnyTran/argilla-data-extractionSignup for coming meetups here: https://lu.ma/d720wy9f
An introduction to distilabel for AI feedback and synthetic data generation Argilla 2023-11-27 | We love the feedback and engagement from our NLP community and their expertise in active learning, LLMs and everything else. We discussed: - the Argilla product vision - argilla-js/ts: a typescript and javascript integration for logging LLM and SLM data back to Argilla- distilabel: AI Feedback framework for building datasets and labelers with LLMs You can find an overview of the shared documents, chat and QnA here: app.session.com/history/HQLEWrDJ9PJ1sOFucQgQmBwAhfn6FpfhiIF3 Colab: colab.research.google.com/drive/1rO1-OlLFPBC0KPuXQOeMpZOeajiwNoMy?usp=sharing Distilabel: github.com/argilla-io/distilabelSignup for coming meetups here: https://lu.ma/d720wy9f
Deploy Argilla on a private Hugging Face space and how to contribute to open source Argilla 2023-11-09 | Another awesome community meetup 🚀We discussed:- deploying Argilla as a private Hugging Face space- the ArgillaTrainer with SetFit and spaCy- how to contribute to open-sourceYou can find an overview of the shared documents, chat and QnA here: app.session.com/history/swdjUegPVToOeC54xp206KRJCK-GczCRPC6AUsed notebook: colab.research.google.com/drive/1WVHmckrjrqGS6jmE-IGAVr8Fbgl840ay?usp=sharingSignup for coming meetups here: https://lu.ma/d720wy9f
Deploy Argilla on a public Hugging Face space and create multi-modal datasets Argilla 2023-10-28 | It was the first Argilla community and we were delighted with the number of attendees and the great engagement!We discussed:- launching Argilla- the new metadata filters- multi-modal supportYou can find an overview of the shared documents, chat and QnA here:app.session.com/history/IYFOuoMSjRDaygbLBPSrV_dCTMlwqr4Q1o38Signup for coming meetups here: https://lu.ma/d720wy9f
Meet Argilla Argilla 2023-09-18 | The feedback layer for enterprise language models: LLMs and SLMs. Build robust language models with human and machine feedback. Argilla empowers data teams from fine-tuning and RLHF to continuous model improvement.
Collect human feedback for evaluating fine-tuned LLMs Argilla 2023-08-28 | In this short video, we show how to collect human feedback for evaluating a fine-tuned LLM (Large Language Model), based on Open AI gpt3.5-turbo, powering a RAG application (Retrieval Augmented Generation).
Collect human feedback for fine-tuning ChatGPT models Argilla 2023-08-28 | In this short video, we show how to collect human feedback for a fine-tuning an LLM (Large Language Model), Open AI gpt3.5-turbo, powering a RAG application (Retrieval Augmented Generation).
Tutorial: Deploy Argilla on Hugging Face and create your first FeebackDataset Argilla 2023-08-23 | Within this video we will showcase how you can use Argilla and our LLM-tailored FeedbackDataset in combination with Hugging Face Spaces and the Hugging Face Hub.Docs: docs.argilla.io/en/latestNotebook: colab.research.google.com/drive/1zTTlfkxyGMujVT98l81w_PoB5_GjuMH6?usp=sharingSlack: join.slack.com/t/rubrixworkspace/shared_invite/zt-20wllqq29-Z11~kp2SeFYjJ0qevJRiPg
Tutorial: Curate an instruction dataset for supervised fine-tuning Argilla 2023-07-04 | Learn how to use Argilla to curate an instruction dataset that you can use to fine-tune an LLM.00:00 Intro01:06 Setup02:53 Define the task09:28 Configure and import your dataset11:12 Annotate in Argilla18:22 Publish your datasetFind a colab notebook with an extended version of this tutorial here: docs.argilla.io/en/latest/guides/llms/examples/curating-feedback-instructiondataset.htmlLearn more about Argilla in...- our documentation: docs.argilla.io/en/latest/index.html- our website: argilla.io
Guía práctica: ¿Cómo mejorar los datos de Alpaca para entrenar un LLM en Español? Argilla 2023-03-28 | En este video tutorial se explican todos los pasos para participar en el esfuerzo colaborativo de mejora de datos para entrenar grandes modelos del lenguaje en Español.Todos los detalles para participar:huggingface.co/datasets/somosnlp/somos-clean-alpaca-es#dataset-card-for-somos-clean-alpaca-esÚnete al Discord de SomosNLP para resolver dudas y compartir tus avances:somosnlp.org¿Qué es Argilla?argilla.io
Argilla: 3 min walkthrough Argilla 2022-10-25 | This video introduces Argilla with a simple example using a pretrained NER model from spaCy to iterate on your own dataset.To reproduce the example you can run:import argilla as rgfrom datasets import load_datasetrecords = rg.read_datasets( load_dataset("argilla/gutenberg_spacy_ner", split="train"), task="TokenClassification",)Based on this tutorial:docs.argilla.io/en/latest/tutorials/notebooks/labelling-tokenclassification-spacy-pretrained.html⭐ Argilla Github repo to stay updated.github.com/argilla/argilla📚 Argilla documentation for more guides and tutorials.docs.argilla.io🙋♀️ Join the Argilla community! A good place to start is our Slackjoin.slack.com/t/rubrixworkspace/shared_invite/zt-whigkyjn-a3IUJLD7gDbTZ0rKlvcJ5g.
