OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Developer Tools & SDKs

Pytorch Transformers Classification

Pytorch Transformers Classification is an open-source Jupyter Notebook repository that provides a starting workflow for training Transformer models on text-classification tasks. It includes code for BERT, XLNet, RoBERTa and XLM, plus notebooks demonstrating data preparation, fine-tuning and evaluation with Yelp Reviews.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Pytorch Transformers Classification — OpenViST project profile in AI Developer Tools & SDKs.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it as an inspectable starting point when studying an older notebook-based training workflow or adapting its examples. Check the library-version caveat before relying on it with a current Transformers installation.

Who it is for

  • Developers studying or adapting a notebook-based Transformer text-classification workflow.
  • Practitioners who want to examine an example using Yelp Reviews and model families including BERT, XLNet, RoBERTa and XLM.

Documented uses

  • Prepare Yelp Reviews files for a text-classification example.
  • Fine-tune a Transformer model on prepared classification data.
  • Evaluate the model in the example workflow.
  • Inspect or adapt training code for BERT, XLNet, RoBERTa or XLM.
Before you choose

Requirements & limits

The repository is public and licensed under Apache-2.0.

  • The README lists BERT, XLNet, RoBERTa and XLM support.
  • The documented setup uses Conda, PyTorch and the older pytorch-transformers package, alongside Jupyter and common data-science dependencies.
  • The README explicitly marks the repository deprecated, says it will not be updated, and warns of possible incompatibility with current Hugging Face Transformers versions.
  • The README directs readers seeking a maintained, simpler training and evaluation library to Simple Transformers.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Prepare the Conda environment and repository

Set up the environment and clone the repository for working with its notebooks.

Before you start

  • Anaconda or Miniconda
  • A machine with Conda available; the README gives separate PyTorch installation instructions for CUDA and CPU

Steps

  1. Create and activate a Conda environment with Python, pandas, tqdm and Jupyter.

  2. Install PyTorch using the README's CUDA or CPU-only instructions.

  3. Install SciPy and scikit-learn with Conda, then install pytorch-transformers and tensorboardX with pip.

  4. Clone the repository using the README's Git command.

Expected outcome

A Conda environment with the listed dependencies and a local copy of the notebook repository.

Guide limitations

  • The repository is deprecated and may not be compatible with current Hugging Face Transformers versions.
  • The README says the project will not be updated.
Documentation 1 ↗
Source-based practical guide

Prepare and run the Yelp classification example

Use the documented Yelp Reviews workflow to prepare data, fine-tune a Transformer model and evaluate results.

Before you start

  • The repository notebooks and dependencies
  • Yelp Reviews dataset files, with train.csv and test.csv placed in the data/ directory

Steps

  1. Download and extract the Yelp Reviews dataset, or use data_download.sh on Linux.

  2. Run data_prep.ipynb to prepare the dataset.

  3. Run run_model.ipynb to fine-tune a Transformer model on the Yelp data and evaluate results.

Expected outcome

The documented workflow prepares Yelp data and uses it to train and evaluate a text-classification model.

Guide limitations

  • The repository is deprecated and may not work with current Hugging Face Transformers versions.
  • The README does not guarantee compatibility with current library releases.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
17 October 2026
Original website title
ThilinaRajapakse/pytorch-transformers-classification · Official homepage title · retained in its source language
How we review sourcesSuggest a correction

Related options

Simple Transformers — The repository README recommends it as the updated, maintained successor for training and evaluating Transformer models with a simpler interface.

Explore project →