AutoTrain

AutoTrain trains custom machine learning models from your own data through a no-code interface. It's for people who need to fine-tune an LLM or build a classifier without writing a training pipeline. The local AutoTrain Advanced project is no longer maintained, so it won't receive bug fixes or new features.

The software is open source under Apache-2.0. You can run training on your own hardware or infrastructure, or use Hugging Face Spaces. The interface also runs on Colab. With the hosted service, training data stays on Hugging Face's servers, private to your account, and transfers use encryption.

AutoTrain covers several kinds of model training:

• LLM fine-tuning, including supervised training, DPO and ORPO preference training, and reward models.
• Text classification, entity recognition, and text regression.
• Translation, summarization, and extractive question answering.
• Image classification and tabular classification or regression.

Automatic model selection is part of the hosted offering: AutoTrain chooses models for the uploaded data, then trains and evaluates them. Its Hugging Face integration makes trained models available on the Hub for serving. Language support follows the models available there.

Training data needs to match the task's expected format. AutoTrain accepts CSV, TSV, JSON, or ZIP files, depending on what you're training.