Source: https://seerepo.com/uses/deepspeed
Markdown source: https://seerepo.com/uses/deepspeed.md
Description: 12 public repos that SeeRepo read use deepspeed. It's used most with torch, transformers, and accelerate.

# Open source repos that use deepspeed

12 public repos that SeeRepo read use deepspeed. It's used most with torch, transformers, and accelerate.

## Used most with

- torch: 5 repos
- transformers: 4 repos
- accelerate: 4 repos
- FastAPI: 3 repos
- sentencepiece: 3 repos
- torchvision: 3 repos
- datasets: 3 repos
- peft: 3 repos

## The repos, by stars

1. [hiyouga/LlamaFactory](https://seerepo.com/r/hiyouga/LlamaFactory): Python · 76,016 stars · 1.6% signed by AI — Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)
2. [ray-project/ray](https://seerepo.com/r/ray-project/ray): Python · 44,282 stars · 3.7% signed by AI — Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
3. [OpenBMB/MiniCPM-V](https://seerepo.com/r/OpenBMB/MiniCPM-V): Python · 27,152 stars · 1.1% signed by AI — See, hear, and interact on your local devices.
4. [liguodongiot/llm-action](https://seerepo.com/r/liguodongiot/llm-action): HTML · 25,281 stars — 本项目旨在分享大模型相关技术原理以及实战经验（大模型工程化、大模型应用落地）
5. [QwenAudio/CosyVoice](https://seerepo.com/r/QwenAudio/CosyVoice): Python · 24,419 stars — Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.
6. [microsoft/unilm](https://seerepo.com/r/microsoft/unilm): Python · 22,240 stars — Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities
7. [BlinkDL/RWKV-LM](https://seerepo.com/r/BlinkDL/RWKV-LM): Python · 14,772 stars — RWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RNN and transformer - great performance, linear time, constant space (no kv-cache), fast training, infinite ctx_len, and free sentence embedding.
8. [FlagOpen/FlagEmbedding](https://seerepo.com/r/FlagOpen/FlagEmbedding): Python · 12,314 stars — Retrieval and Retrieval-augmented LLMs
9. [OpenRLHF/OpenRLHF](https://seerepo.com/r/OpenRLHF/OpenRLHF): Python · 10,152 stars · 1.8% signed by AI — An Easy-to-use, Scalable and High-performance Agentic RL Framework based on Ray (PPO & DAPO & REINFORCE++ & VLM & TIS & vLLM & Ray & Async RL)
10. [modelscope/modelscope](https://seerepo.com/r/modelscope/modelscope): Python · 9,218 stars · 0.1% signed by AI — ModelScope: bring the notion of Model-as-a-Service to life.
11. [apple-aiml-research/ml-ferret](https://seerepo.com/r/apple-aiml-research/ml-ferret): Python · 8,646 stars
12. [OpenBMB/XAgent](https://seerepo.com/r/OpenBMB/XAgent): Python · 8,549 stars · 1.1% signed by AI — An Autonomous LLM Agent for Complex Task Solving

## Frequently asked questions

### What is the most popular open source repo that uses deepspeed?

hiyouga/LlamaFactory, with 76,016 stars on GitHub.

The proof: [hiyouga/LlamaFactory on GitHub](https://github.com/hiyouga/LlamaFactory)

### What other technologies is deepspeed used with?

With torch (5), transformers (4), accelerate (4), FastAPI (3), and sentencepiece (3), out of the 12 repos that use it.

### What languages are the repos that use deepspeed written in?

Python (11) and HTML (1).

### How many repos that use deepspeed have commits signed by AI?

6 of the 12; the most, ray-project/ray, with 3.7% of its commits.

The proof: [the commits each tool signed in ray-project/ray](https://seerepo.com/r/ray-project/ray/ai)

## What is your repo built with?

Any public repo, with no account: https://seerepo.com/r/owner/repo/built, with the owner and name it has on GitHub.
