Open Source GPU & Compute Platforms Software

GPU & Compute Platforms provide environments for training, evaluating, and scaling large language and diffusion models. These tools manage GPU provisioning and container orchestration across local hardware and clusters. This type of software serves developers and researchers who need to run inference or fine-tune models on their own infrastructure. Self-hosting these apps allows you to manage hardware allocation and model weights without relying on external cloud providers. You can coordinate compute resources and distribute workloads across your own servers to avoid external API constraints.

This page lists 2 open source tools in the GPU & Compute Platforms category. The most popular are Transformer Lab and dstack. Most use the AGPL-3.0 or MPL-2.0 license, and 2 offer an official Docker image.