OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

C Tran

C Tran is an MIT-licensed Python model and codebase for transformer-based multi-label image classification, with documented training paths for COCO80 and VOC20.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

C Tran — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use C Tran when you want to train the model on one of its documented image datasets and can work with its Python setup and dataset archives.

Who it is for

  • Computer-vision practitioners training multi-label image classifiers on COCO80 or VOC20.
  • Developers exploring transformer-based methods for assigning multiple labels to images.

Documented uses

  • Train a multi-label image classifier on COCO80.
  • Train a multi-label image classifier on VOC20.
Before you choose

Requirements & limits

The code is publicly available under the MIT license.

  • Python 3.7 is required, and package versions are listed in requirements.txt.
  • The README lists the COCO archive as 19 GB and the VOC2007 archive as 1.7 GB.
  • The repository provides separate training commands and dataset flags for COCO80 and VOC20.

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

Train on COCO80

Prepare the documented COCO data and start C Tran training with the repository's COCO command.

Before you start

  • Python 3.7
  • Packages and versions listed in requirements.txt
  • The COCO dataset archive, listed as 19 GB

Steps

  1. Download the COCO archive from the URL in the repository README.

  2. Create a data/ directory and extract the archive into it.

  3. Run the README's Python training command with --dataset 'coco' and --dataroot data/.

Expected outcome

The documented command starts training a new C Tran model on COCO80.

Guide limitations

  • The README documents this training path for COCO80; it does not provide instructions here for adapting to other datasets.
Documentation 1 ↗
Source-based practical guide

Train on VOC20

Prepare the documented VOC2007 data and start C Tran training with the repository's VOC command.

Before you start

  • Python 3.7
  • Packages and versions listed in requirements.txt
  • The VOC2007 archive, listed as 1.7 GB

Steps

  1. Download the VOC archive from the URL in the repository README.

  2. Create a data/ directory and extract the archive into it.

  3. Run the README's Python training command with --dataset 'voc', --grad_ac_step 2, and --dataroot data/.

Expected outcome

The documented command starts training a new C Tran model on VOC20.

Guide limitations

  • The README documents this training path for VOC20; it does not provide instructions here for adapting to other datasets.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
QData/C-Tran · Official homepage title · retained in its source language
How we review sourcesSuggest a correction