Training Transformers Together

This organization is a part of the NeurIPS 2021 demonstration "Training Transformers Together".

In this demo, we've trained a model similar to OpenAI DALL-E — a Transformer "language model" that generates images from text descriptions. Training happened collaboratively — volunteers from all over the Internet contributed to the training using hardware available to them. We used LAION-400M, the world's largest openly available image-text-pair dataset with 400 million samples. Our model was based on the dalle‑pytorch implementation by Phil Wang with a few tweaks to make it communication-efficient.

See details about how it works on our website.

This organization gathers people participating in the collaborative training and provides links to the related materials:

• 👉 Inference results
• 👉 Model weights (the latest checkpoint)
• 👉 Colab notebook for running inference (updated on Apr 5)
• 👉 Code
• 👉 Dataset

The materials below were available during the training run itself:

• 👉 Starter kits for Google Colab and Kaggle (easy way to join the training)
• 👉 Dashboard (the current training state: loss, number of peers, etc.)
• 👉 Weights & Biases plots for aux peers (aggregating the metrics) and actual trainers (contributing with their GPUs)

Feel free to reach us on Discord if you have any questions 🙂