Kubeflow Trainer

What it does

Community support

The project links users to a Kubeflow Trainer Slack channel and regular community calls.checked 11 Oct 2026

Data cache

Its distributed data cache uses Apache Arrow and Apache DataFusion for zero-copy tensor streaming to GPU nodes.checked 11 Oct 2026

Deployment environments

The overview says it can run in public cloud, on-premises, or hybrid environments.checked 11 Oct 2026

Fine-tuning

Built-in TorchTune workflows support LoRA, QLoRA, and full fine-tuning with HuggingFace model and dataset URIs.checked 11 Oct 2026

Frameworks

It provides a unified Python SDK and TrainJob API for frameworks including PyTorch, JAX, DeepSpeed, MLX, HuggingFace, Megatron, and XGBoost.checked 11 Oct 2026

Installation requirement

The installation guide lists Kubernetes >= 1.31 and kubectl >= 1.31 as minimum requirements for the control plane.checked 11 Oct 2026

Integrations

The project documents integrations with Kueue, Slurm Bridge, KAI Scheduler, JobSet, and LeaderWorkerSet.checked 11 Oct 2026

License

The public GitHub repository lists an Apache-2.0 license.checked 11 Oct 2026

Local execution

TrainJobs can run locally with Docker or Podman and then deploy to Kubernetes environments without code changes.checked 11 Oct 2026

Project status

The GitHub repository README states that Kubeflow Trainer is currently in alpha status and its APIs may change.checked 11 Oct 2026

Purpose

Kubeflow Trainer is a Kubernetes-native platform for distributed AI model training and LLM fine-tuning.checked 11 Oct 2026

Scale

It supports training from a single GPU to multi-node clusters, with automatic setup for DDP, FSDP, parameter servers, and gang scheduling.checked 11 Oct 2026

Supported users

The documentation is organized for AI practitioners, platform administrators, and open-source contributors.checked 11 Oct 2026