Source: https://seerepo.com/r/hiyouga/LlamaFactory/built
Markdown source: https://seerepo.com/r/hiyouga/LlamaFactory/built.md
Description: LlamaFactory is a Python repo built with FastAPI and uvicorn. It uses AI through einops, sentencepiece, tiktoken, torchvision, triton-ascend, and 14 more.

# What is LlamaFactory built with?

LlamaFactory is a Python repo built with FastAPI and uvicorn. It uses AI through einops, sentencepiece, tiktoken, torchvision, triton-ascend, and 14 more, makes its content with torchcodec and av, validates its data with jsonschema-specifications, referencing, and Pydantic, and is tested with Pytest. It has 62,624 written lines, with 62 test files.

Read on November 4, 2026, at commit 702005c.

Repository: https://github.com/hiyouga/LlamaFactory

## Frontend

### Interface

- **Gradio**: 14 files
- **[matplotlib](https://seerepo.com/uses/matplotlib)**: 1 file

## Backend

### The base

- **[FastAPI](https://seerepo.com/uses/fastapi)**: 3 files

### AI

- **einops**
- **sentencepiece**
- **[tiktoken](https://seerepo.com/uses/tiktoken)**
- **torchvision**
- **triton-ascend**
- **[transformers](https://seerepo.com/uses/transformers)**: 76 files
- **[torch](https://seerepo.com/uses/torch)**: 73 files
- **datasets**: 12 files
- **accelerate**: 11 files
- **trl**: 11 files
- **peft**: 9 files
- **deepspeed**: 4 files
- **safetensors**: 4 files
- **liger-kernel**: 2 files
- **ModelScope**: 2 files
- **torchaudio**: 2 files
- **bitsandbytes**: 1 file
- **nltk**: 1 file
- **torchdata**: 1 file

## Support

### The base

- **[uvicorn](https://seerepo.com/uses/uvicorn)**: 2 files

### Content

- **torchcodec**
- **av**: 1 file

### Validation

- **jsonschema-specifications**
- **referencing**
- **[Pydantic](https://seerepo.com/uses/pydantic)**: 2 files

### Testing

- **Pytest**: 49 files

### Tools

- **build**
- **fire**
- **pre-commit**
- **Ruff**
- **tyro**
- **packaging**: 3 files

### Other

- **hf-transfer**
- **pandas**
- **protobuf**
- **SciPy**
- **soundfile**
- **NumPy**: 13 files
- **omegaconf**: 4 files
- **pyyaml**: 2 files
- **jieba**: 1 file
- **sse-starlette**: 1 file

## Other repos with a similar stack

- [unslothai/unsloth](https://seerepo.com/r/unslothai/unsloth/built): FastAPI, uvicorn, matplotlib, einops, sentencepiece, tiktoken, transformers, torch, datasets, accelerate, peft, safetensors, av, Pydantic · [Compare](https://seerepo.com/compare/hiyouga/LlamaFactory/vs/unslothai/unsloth)
- [vllm-project/vllm](https://seerepo.com/r/vllm-project/vllm/built): FastAPI, einops, sentencepiece, tiktoken, torchvision, transformers, torch, datasets, peft, safetensors, torchaudio, Pydantic · [Compare](https://seerepo.com/compare/hiyouga/LlamaFactory/vs/vllm-project/vllm)
- [datawhalechina/hello-agents](https://seerepo.com/r/datawhalechina/hello-agents/built): FastAPI, uvicorn, Gradio, matplotlib, tiktoken, transformers, torch, datasets, accelerate, peft, safetensors, Pydantic · [Compare](https://seerepo.com/compare/datawhalechina/hello-agents/vs/hiyouga/LlamaFactory)
- [debpalash/VoiceStudio](https://seerepo.com/r/debpalash/VoiceStudio/built): FastAPI, Gradio, matplotlib, torchvision, transformers, torch, accelerate, ModelScope, torchaudio, Pydantic · [Compare](https://seerepo.com/compare/debpalash/VoiceStudio/vs/hiyouga/LlamaFactory)
- [comet-ml/opik](https://seerepo.com/r/comet-ml/opik/built): FastAPI, tiktoken, transformers, torch, datasets, accelerate, peft, referencing, Pydantic · [Compare](https://seerepo.com/compare/comet-ml/opik/vs/hiyouga/LlamaFactory)
- [rohitg00/ai-engineering-from-scratch](https://seerepo.com/r/rohitg00/ai-engineering-from-scratch/built): matplotlib, tiktoken, torchvision, transformers, torch, datasets, accelerate, torchaudio · [Compare](https://seerepo.com/compare/hiyouga/LlamaFactory/vs/rohitg00/ai-engineering-from-scratch)

## Frequently asked questions

### What language is LlamaFactory written in?

LlamaFactory is written in Python, with 62,624 written lines.

### What framework does LlamaFactory use?

LlamaFactory is built with FastAPI and uvicorn.

### Does LlamaFactory use AI?

Yes. LlamaFactory uses AI through einops, sentencepiece, tiktoken, torchvision, triton-ascend, transformers, torch, datasets, accelerate, trl, peft, deepspeed, safetensors, liger-kernel, ModelScope, torchaudio, bitsandbytes, nltk, and torchdata.

### Does LlamaFactory have tests?

Yes. LlamaFactory has 62 test files and is tested with Pytest.

### Which repos have a stack like LlamaFactory's?

The ones that share most of its stack: unslothai/unsloth, vllm-project/vllm, and datawhalechina/hello-agents.

## 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.