Skip to main content
Generative AI

Together, Cloudflare and MosaicML give users the freedom to train LLMs on any compute, anywhere in the world, for faster, cheaper training runs without vendor lock-in.

Read the complete blog post to learn more!

Building generative AI models requires massive compute AND data storage infrastructure. Training huge datasets means that terabytes of data must be read in parallel by thousands of processes. In addition, model checkpoints need to be saved frequently throughout a training run, and these checkpoints alone can be hundreds of gigabytes in size.

In a recent blog post, Cloudflare and MosaicML engineers discuss how their tools work together to address these challenges. MosaicML’s open source StreamingDataset and Composer libraries let users easily stream in training data and read/write model checkpoints back to Cloudflare R2. And thanks to R2’s zero-egress pricing and MosaicML’s cloud-agnostic platform, users can start/stop/move/resize jobs in response to GPU availability and prices across compute providers, without paying any data transfer fees. By eliminating egress fees, R2’s storage is an exceptionally cost-effective complement to MosaicML training, providing maximum autonomy and control.

Try Databricks for free

Related posts

Generative AI

MosaicML StreamingDataset: Fast, Accurate Streaming of Training Data from Cloud Storage

Loading your training data becomes an escalating challenge as datasets grow bigger in size and the number of nodes scales. We built StreamingDataset...
Generative AI

Announcing MPT-7B-8K: 8K Context Length for Document Understanding

July 18, 2023 by Sam Havens and Erica Ji Yuen in Generative AI
Today, we are releasing MPT-7B-8K, a 7B parameter open-source LLM with 8k context length trained with the MosaicML platform. MPT-7B-8K was pretrained starting...
Generative AI

MPT-30B: Raising the bar for open-source foundation models

Introducing MPT-30B, a new, more powerful member of our Foundation Series of open-source models, trained with an 8k context length on NVIDIA H100...
See all Generative AI posts