RepoRadar
Leave a Tip
Item detail

oumi-ai/oumi

Oumi oumi-ai/oumi is an Apache-2.0 fully open-source AI platform from Oumi that covers the entire foundation-model lifecycle in one repository, including training and fine-tuning (full-parameter SFT, LoRA, QLoRA, DPO, RLHF, distributed training on multi-GPU and multi-node), evaluation (50+ academic and custom benchmarks including MMLU, GSM8K, AlpacaEval, and a built-in judge model), data preparati

Open source →Back to radar
Score7.6
Popularity9319.0
Risklow
TierGold
Score breakdown
Usefulness8.0
Novelty10.0
Momentum10.0
Maturity8.5
Open-source/build7.4
Evidence7.2
Workflow potential8.3
Setup ease6.5

Popularity is tracked separately. Support, ads, sponsorships, and tips never affect these signals.

Why it matters

Useful for AI engineers, ML researchers, and platform teams who want a single Apache-2.0 platform to train, fine-tune, evaluate, and deploy open foundation models across multiple backends (Hugging Face, llama.cpp, vLLM, SGLang) instead of stitching together four separate frameworks, because oumi-ai/oumi ships training (full-parameter SFT, LoRA, QLoRA, DPO, distributed multi-GPU / multi-node), eval

Who should use it

BuildersPower users

Who should skip it

Skip if the source link, docs, or setup requirements do not match your workflow.

Risk explanation

Risk label needs manual review.

Evidence links

Closest alternatives / related signals