Today's Fine-tuning & Training: Fastest-Growing Projects — May 09, 2026
Today's the Fine-tuning & Training space, we've seen a surge in interest around tools that enable efficient and effective training of large language models (LLMs). Many repositories are gaining traction by providing innovative solutions for fine-tuning and adapting LLMs to specific tasks or languages. As researchers and developers continue to push the boundaries of what's possible with AI, these tools are becoming increasingly important.
QingGo/engram-peft has taken the top spot this week with a growth score of 12.04 and 32 stars. This unofficial implementation of DeepSeek Engram allows users to inject high-capacity conditional memory into LLMs via sparse retrieval PEFT without increasing inference FLOPs, making it an attractive solution for those looking to improve model performance without sacrificing efficiency. Its rapid growth is likely due to the increasing demand for more advanced and efficient fine-tuning techniques.
UNfukashigi/Anima-LoRA-Factory has seen significant growth with a score of 7.50 and 27 stars. This user-friendly GUI tool is designed specifically for training LoRAs for next-generation Anima diffusion models, making it an essential resource for researchers working in this area. Its popularity can be attributed to the growing interest in Anima models and the need for intuitive tools to support their development.
ZJU-OmniAI/GFT boasts a growth score of 6.65 and 30 stars. This repository provides a novel approach to fine-tuning, leveraging unbiased group advantages and dynamic coefficient rectification to achieve better results. Its growth is likely driven by researchers seeking new methods to improve the performance of their models.
semidark/kokoro-deutsch has gained traction with a growth score of 4.98 and 34 stars. This project offers a complete training recipe for fine-tuning Kokoro-82M on German, providing a valuable resource for those looking to adapt this model to new languages. Its popularity can be attributed to the increasing demand for multilingual support in AI models.
raiyanyahya/how-to-train-your-gpt has seen steady growth with a score of 4.69 and an impressive 774 stars. This repository provides a comprehensive guide to building a modern LLM from scratch, making it an invaluable resource for newcomers to the field. Its enduring popularity is likely due to its accessibility and the growing interest in building custom AI models.
Jackohhhh/MedLLM-Finetuning has achieved a growth score of 3.20 and 21 stars. This out-of-the-box LLM fine-tuning framework for medical binary classification tasks offers a convenient solution for researchers working in this area. Its popularity can be attributed to the growing demand for specialized AI models in the healthcare sector.
Mintzs/oogaboogalm has seen moderate growth with a score of 2.10 and 47 stars. This repository explores the idea of baking caveman system prompts and skills into LLMs through fine-tuning, offering an innovative approach to reducing token use. Its growth is likely driven by researchers seeking novel methods to improve model efficiency.
hlpun/Train-in-Silence has gained some traction with a growth score of 1.37 and 44 stars. This Task-Aware MCP server and automated VRAM calculator for LLM fine-tuning provides a useful tool for optimizing training processes. Its popularity can be attributed to the increasing demand for efficient training solutions in the AI community.
Note that we've skipped repositories with no meaningful description, focusing on those that provide valuable insights into the trends and innovations in the Fine-tuning & Training space.
QingGo/engram-peft has taken the top spot this week with a growth score of 12.04 and 32 stars. This unofficial implementation of DeepSeek Engram allows users to inject high-capacity conditional memory into LLMs via sparse retrieval PEFT without increasing inference FLOPs, making it an attractive solution for those looking to improve model performance without sacrificing efficiency. Its rapid growth is likely due to the increasing demand for more advanced and efficient fine-tuning techniques.
UNfukashigi/Anima-LoRA-Factory has seen significant growth with a score of 7.50 and 27 stars. This user-friendly GUI tool is designed specifically for training LoRAs for next-generation Anima diffusion models, making it an essential resource for researchers working in this area. Its popularity can be attributed to the growing interest in Anima models and the need for intuitive tools to support their development.
ZJU-OmniAI/GFT boasts a growth score of 6.65 and 30 stars. This repository provides a novel approach to fine-tuning, leveraging unbiased group advantages and dynamic coefficient rectification to achieve better results. Its growth is likely driven by researchers seeking new methods to improve the performance of their models.
semidark/kokoro-deutsch has gained traction with a growth score of 4.98 and 34 stars. This project offers a complete training recipe for fine-tuning Kokoro-82M on German, providing a valuable resource for those looking to adapt this model to new languages. Its popularity can be attributed to the increasing demand for multilingual support in AI models.
raiyanyahya/how-to-train-your-gpt has seen steady growth with a score of 4.69 and an impressive 774 stars. This repository provides a comprehensive guide to building a modern LLM from scratch, making it an invaluable resource for newcomers to the field. Its enduring popularity is likely due to its accessibility and the growing interest in building custom AI models.
Jackohhhh/MedLLM-Finetuning has achieved a growth score of 3.20 and 21 stars. This out-of-the-box LLM fine-tuning framework for medical binary classification tasks offers a convenient solution for researchers working in this area. Its popularity can be attributed to the growing demand for specialized AI models in the healthcare sector.
Mintzs/oogaboogalm has seen moderate growth with a score of 2.10 and 47 stars. This repository explores the idea of baking caveman system prompts and skills into LLMs through fine-tuning, offering an innovative approach to reducing token use. Its growth is likely driven by researchers seeking novel methods to improve model efficiency.
hlpun/Train-in-Silence has gained some traction with a growth score of 1.37 and 44 stars. This Task-Aware MCP server and automated VRAM calculator for LLM fine-tuning provides a useful tool for optimizing training processes. Its popularity can be attributed to the increasing demand for efficient training solutions in the AI community.
Note that we've skipped repositories with no meaningful description, focusing on those that provide valuable insights into the trends and innovations in the Fine-tuning & Training space.