Torchtune
by PyTorch / Meta · open-source · Last verified 2026-04-24
Torchtune is PyTorch's official fine-tuning library for LLMs, providing modular, easy-to-customize training recipes for LoRA, QLoRA, and full fine-tuning. Built with composability in mind, it prioritizes readability over abstraction so practitioners can understand and modify every training component. Supports Llama, Mistral, Gemma, and Phi model families.
https://github.com/pytorch/torchtune ↗C
C—Below Average
Adoption: C+Quality: B+Freshness: ACitations: CEngagement: F
Specifications
- License
- Open Source
- Pricing
- open-source
- Capabilities
- Integrations
- Use Cases
- API Available
- No
- SDK Languages
- Tags
- fine-tuning, pytorch, official, lora, qlora, modular, readable
- Added
- 2026-04-24
- Completeness
- 60%
Index Score
44Adoption
50
Quality
70
Freshness
80
Citations
40
Engagement
0