← OSS.Radar home

Powered by aegismemory.com · Aegis Memory repository

Fine Tuning AI repositories

OSS Radar projects in the fine tuning category.

rasbt/LLMs-from-scratch

Implement a ChatGPT-like LLM in PyTorch from scratch, step by step

Category
fine tuning
Stars
100,964
Readiness
needs review (67/100 heuristic points; not a probability)
Data confidence
medium
Maintainer health
healthy
Maintenance risk
0/100 · high confidence

Why: +734 stars in 7 days; 63 lifetime contributors

Why it may be a gem: healthy maintenance and project fundamentals; open issue backlog is stable or shrinking

Strongest signals: push recency, contributor breadth, issue load. Risks: maintenance is concentrated in one contributor. Missing inputs: release recency.

ai artificial-intelligence attention-mechanism deep-learning finetuning from-scratch

hiyouga/LlamaFactory

Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)

Category
fine tuning
Stars
73,903
Readiness
ready (83/100 heuristic points; not a probability)
Data confidence
high
Maintainer health
healthy
Maintenance risk
0/100 · high confidence

Why: +252 stars in 7 days; 17 commits in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; consistent human and community activity

Strongest signals: push recency, contributor breadth, issue load. Risks: None identified. Missing inputs: None.

Capped lower bounds: lifetime contributors.

agent ai deepseek fine-tuning gemma gpt

datawhalechina/self-llm

《开源大模型食用指南》针对中国宝宝量身打造的基于Linux环境快速微调（全参数/Lora）、部署国内外开源大模型（LLM）/多模态大模型（MLLM）教程

Category
fine tuning
Stars
31,610
Readiness
needs review (68/100 heuristic points; not a probability)
Data confidence
medium
Maintainer health
healthy
Maintenance risk
34/100 · high confidence

Why: +105 stars in 7 days; 79 lifetime contributors

Why it may be a gem: open issue backlog is stable or shrinking

Strongest signals: push recency, contributor breadth, issue load. Risks: 163 open issues with no maintainer issue responses in 30 days, no pull-request review responses in 30 days, no maintainer response activity in 30 days. Missing inputs: release recency.

chatglm chatglm3 gemma-2b-it glm-4 internlm2 llama3

lyogavin/airllm

AirLLM 70B inference with single 4GB GPU

Category
fine tuning
Stars
29,913
Readiness
ready (92/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +5,566 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chinese-llm chinese-nlp finetune generative-ai instruct-gpt instruction-set

ymcui/Chinese-LLaMA-Alpaca

中文LLaMA&Alpaca大语言模型+本地CPU/GPU训练部署 (Chinese LLaMA & Alpaca LLMs)

Category
fine tuning
Stars
18,942
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
19/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 111 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

alpaca alpaca-2 large-language-models llama llama-2 llm

ConardLi/easy-dataset

A powerful tool for creating datasets for LLM fine-tuning 、RAG and Eval

Category
fine tuning
Stars
14,760
Readiness
needs review (46/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
16/100 · low confidence

Why: +25 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 98 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

dataset fine-tuning javascript llm rag

RUCAIBox/LLMSurvey

The official GitHub page for the survey paper "A Survey of Large Language Models".

Category
fine tuning
Stars
12,203
Readiness
high risk (41/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +7 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 514 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chain-of-thought chatgpt in-context-learning instruction-tuning large-language-models llm

yangjianxin1/Firefly

Firefly: 大模型训练工具，支持训练Qwen2.5、Qwen2、Yi1.5、Phi-3、Llama3、Gemma、MiniCPM、Yi、Deepseek、Orion、Xverse、Mixtral-8x7B、Zephyr、Mistral、Baichuan2、Llma2、Llama、Qwen、Baichuan、ChatGLM2、InternLM、Ziya2、Vicuna、Bloom等大模型

Category
fine tuning
Stars
6,651
Readiness
high risk (41/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 652 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

alpaca aquila baichuan chatglm gemma gpt

smallcloudai/refact

AI Agent that handles engineering tasks end-to-end: integrates with developers’ tools, plans, executes, and iterates until it achieves a successful result.

Category
fine tuning
Stars
3,544
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
70/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: repository is archived. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-agent developer-tools enterprise fine-tuning on-prem open-source

aiming-lab/MetaClaw

🦞 Just talk to your agent — it learns and EVOLVES 🧬.

Category
fine tuning
Stars
3,518
Readiness
needs review (56/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +16 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent ai-agent continual-learning fine-tuning llm lora

SakanaAI/doc-to-lora

Hypernetworks that update LLMs to remember factual information

Category
fine tuning
Stars
794
Readiness
needs review (60/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai ai-agent hypernetworks llm llm-agent lora

decodingai-magazine/second-brain-ai-assistant-course

Learn to build your Second Brain AI assistant with LLMs, agents, RAG, fine-tuning, LLMOps and AI systems techniques.

Category
fine tuning
Stars
2,955
Readiness
needs review (58/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
0/100 · low confidence

Why: +15 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agents ai-systems data-engineering fine-tuning huggingface llm

iusztinpaul/hands-on-llms

🦖 𝗟𝗲𝗮𝗿𝗻 about 𝗟𝗟𝗠𝘀, 𝗟𝗟𝗠𝗢𝗽𝘀, and 𝘃𝗲𝗰𝘁𝗼𝗿 𝗗𝗕𝘀 for free by designing, training, and deploying a real-time financial advisor LLM system ~ 𝘴𝘰𝘶𝘳𝘤𝘦 𝘤𝘰𝘥𝘦 + 𝘷𝘪𝘥𝘦𝘰 & 𝘳𝘦𝘢𝘥𝘪𝘯𝘨 𝘮𝘢𝘵𝘦𝘳𝘪𝘢𝘭𝘴

Category
fine tuning
Stars
3,423
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: repository is archived, no push in 606 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

3-pipeline-design aws beam bytewax cicd comet-ml

Oneirocom/Magick

Magick is a cutting-edge toolkit for a new kind of AI builder. Make Magick with us!

Category
fine tuning
Stars
846
Readiness
needs review (49/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 409 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent agi ai blueprints-visual-scripting embeddings fine-tuning

huggingface/peft

🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.

Category
fine tuning
Stars
21,515
Readiness
ready (92/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +44 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

adapter diffusion fine-tuning llm lora parameter-efficient-learning

axolotl-ai-cloud/axolotl

Go ahead and axolotl questions

Category
fine tuning
Stars
12,326
Readiness
ready (86/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +35 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning llm

ludwig-ai/ludwig

Low-code framework for building custom LLMs, neural networks, and other AI models

Category
fine tuning
Stars
11,749
Readiness
ready (90/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +4 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

computer-vision data-centric data-science deep deep-learning deeplearning

oumi-ai/oumi

Easily fine-tune, evaluate and deploy Qwen, Gemma, or any open weight LLM!

Category
fine tuning
Stars
9,368
Readiness
ready (91/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

dpo evaluation fine-tuning gpt-oss gpt-oss-120b gpt-oss-20b

flyteorg/flyte

Dynamic, resilient AI orchestration. Coordinate data, models, and compute as you build AI workflows.

Category
fine tuning
Stars
7,182
Readiness
ready (93/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +13 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agentic ai-agents ai-development-tools data-analysis data-science declarative

h2oai/h2o-llmstudio

H2O LLM Studio - a framework and no-code GUI for fine-tuning LLMs. Documentation: https://docs.h2o.ai/h2o-llmstudio/

Category
fine tuning
Stars
5,047
Readiness
ready (91/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai chatbot chatgpt fedramp fine-tuning finetuning

Kiln-AI/Kiln

Build, Evaluate, and Optimize AI Systems. Includes evals, RAG, agents, fine-tuning, synthetic data generation, dataset management, MCP, and more.

Category
fine tuning
Stars
5,012
Readiness
ready (82/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +13 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai chain-of-thought collaboration dataset-generation evals evaluation

AIDotNet/OpenDeepWiki

OpenDeepWiki is the open-source version of the DeepWiki project, aiming to provide a powerful knowledge management and collaboration platform. The project is mainly developed using C# and TypeScript,

Category
fine tuning
Stars
3,521
Readiness
ready (91/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +55 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chatgpt deepwiki docs fine-tuning

Nerogar/OneTrainer

OneTrainer is a one-stop solution for all your Diffusion training needs.

Category
fine tuning
Stars
3,153
Readiness
ready (80/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +9 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning image-model-training lora training

bghira/SimpleTuner

A general fine-tuning kit geared toward image/video/audio diffusion models.

Category
fine tuning
Stars
2,889
Readiness
ready (83/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

diffusers diffusion-models fine-tuning flux-dev machine-learning stable-diffusion

roboflow/maestro

streamline the fine-tuning process for multimodal models: PaliGemma 2, Florence-2, and Qwen2.5-VL

Category
fine tuning
Stars
2,691
Readiness
ready (88/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +10 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

captioning fine-tuning florence-2 multimodal objectdetection paligemma

pykeio/ort

Fast ML inference & training for ONNX models in Rust

Category
fine tuning
Stars
2,444
Readiness
ready (92/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +16 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai ai-training fine-tuning inference machine-learning onnx

AI-Hypercomputer/maxtext

A simple, performant, and scalable Jax LLM!

Category
fine tuning
Stars
2,381
Readiness
ready (98/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +5 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

deepseek fine-tuning gemma2 gemma3 gpt jax

dstackai/dstack

Vendor-agnostic orchestration for training, inference and agentic workloads across NVIDIA, AMD, TPU, and Tenstorrent on clouds, Kubernetes, and bare metal.

Category
fine tuning
Stars
2,208
Readiness
ready (84/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +6 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent-skills agentic-orchestration amd cloud containers docker

kubeflow/trainer

Distributed AI Model Training and LLM Fine-Tuning on Kubernetes

Category
fine tuning
Stars
2,174
Readiness
ready (99/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +9 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai distributed fine-tuning gpu huggingface jax

NVIDIA-NeMo/Nemotron

Developer Asset Hub for NVIDIA Nemotron — A one-stop resource for training recipes, usage cookbooks, datasets, and full end-to-end reference examples to build with Nemotron models

Category
fine tuning
Stars
1,830
Readiness
ready (94/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +29 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai fine-tuning model-training nemotron nvidia reinforcement-learning

beam-cloud/beta9

Ultrafast serverless GPU inference, sandboxes, and background jobs

Category
fine tuning
Stars
1,732
Readiness
ready (83/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +6 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

autoscaler cloudrun cuda developer-productivity distributed-computing faas

bespokelabsai/curator

Synthetic data curation for post-training and structured data extraction

Category
fine tuning
Stars
1,711
Readiness
ready (90/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +5 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agents deep-learning fine-tuning instruction-tuning llm machine-learning

NVIDIA-NeMo/Curator

Scalable data pre processing and curation toolkit for LLMs

Category
fine tuning
Stars
1,703
Readiness
ready (93/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +9 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

data data-curation data-prep data-preparation data-processing data-processing-pipelines

R6410418/Jackrong-llm-finetuning-guide

No description

Category
fine tuning
Stars
1,623
Readiness
ready (78/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +8 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

dataset deepseek fine-tuning guide llama3 llm

nolabs-ai/deepfabric

Generate High-Quality Synthetics, Train, Measure, and Evaluate in a Single Pipeline

Category
fine tuning
Stars
880
Readiness
ready (86/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agents ai data-science dataset distillation evaluation

invergent-ai/surogate

Training/Fine-tuning at the speed of light

Category
fine tuning
Stars
807
Readiness
ready (87/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

cuda deep-learning fine-tuning generative-ai llama llm

DaoyuanLi2816/can-i-finetune-this

Estimate whether a Hugging Face model fits and fine-tunes on your local GPU.

Category
fine tuning
Stars
792
Readiness
ready (84/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bitsandbytes fine-tuning gpu hugging-face llm lora

kaito-project/aikit

🏗️ Fine-tune, build, and deploy open-source LLMs easily!

Category
fine tuning
Stars
535
Readiness
ready (91/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai buildkit chatgpt docker fine-tuning finetuning

Tavris1/AI-Toolkit-Easy-Install

One-click Portable Windows installation of 'AI-Toolkit by Ostris'

Category
fine tuning
Stars
499
Readiness
needs review (63/100 heuristic points; not a probability)
Data confidence
high
Maintainer health
watch
Maintenance risk
0/100 · high confidence

Why: +5 stars in 7 days; 11 commits in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; open issue backlog is stable or shrinking

Strongest signals: push recency, issue load, documentation. Risks: maintenance is concentrated in one contributor. Missing inputs: None.

ai-toolkit easy-install fine-tuning lora machine-learning stable-diffusion

open-edge-platform/geti_v2

⚠️ Legacy repository for Geti v2.x. For Geti v3.0+, visit https://github.com/open-edge-platform/geti

Category
fine tuning
Stars
484
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
high
Maintainer health
risky
Maintenance risk
88/100 · high confidence

Why: 36 lifetime contributors

Why it may be a gem: healthy maintenance and project fundamentals; open issue backlog is stable or shrinking

Strongest signals: push recency, contributor breadth, issue load. Risks: repository is archived, recent commit cadence is 2/10.6 of its monthly baseline. Missing inputs: None.

computer-vision deep-learning fine-tuning geti inference openvino

MakazhanAlpamys/Soup

Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU.

Category
fine tuning
Stars
398
Readiness
ready (90/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +317 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

cli consumer-gpu dpo fine-tuning gguf huggingface

ai4protein/VenusFactory2

🏭 AI agent platform with skills for protein engineering, the noob-friendly AI tutorial tool for life science professionals.

Category
fine tuning
Stars
249
Readiness
ready (79/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent database fine-tuning language-model life-science noob-friendly

huggingface/optimum-habana

Easy and lightning fast training of 🤗 Transformers on Habana Gaudi processor (HPU)

Category
fine tuning
Stars
212
Readiness
ready (94/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bert fine-tuning habana hpu transformers

Enping-Hu/minimind-deep-dive

从 MiniMind 源码读起,再延伸到现代大模型技术体系的中文学习笔记。主线逐行精读预训练 / SFT / DPO / PPO / GRPO 与训练机制;附录 17 篇进阶卷覆盖量化、投机解码、RLHF 全景、模型代际史等 MiniMind 没涉及、但进阶绕不开的主题。

Category
fine tuning
Stars
192
Readiness
needs review (71/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

deep-learning dpo fine-tuning from-scratch grpo large-language-models

RapidFireAI/rapidfireai

RapidFire AI: Rapid AI Customization from RAG to Fine-Tuning

Category
fine tuning
Stars
167
Readiness
ready (89/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai artifical-intelligense context-engineering deep-learning experiment-tracking experimentation

FennelFetish/qapyq

AI-assisted media curator for large image/video datasets. Streamlined captioning, cropping, masking for LoRA/diffusion training workflows.

Category
fine tuning
Stars
163
Readiness
ready (80/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai annotation automation desktop-app fine-tuning image

philfung/awesome-reliable-robotics

Robotics research demonstrating reliability and robustness in the real world (continuously updated)

Category
fine tuning
Stars
161
Readiness
ready (79/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning imitation-learning manipulation manipulator-robotics reinforcement-learning reward-shaping

Zyora-Dev/zse

The inference engine the open-source world built for itself.

Category
fine tuning
Stars
153
Readiness
ready (72/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai amd-gpu apple apple-silicon fine-tuning inference-engine

riyanshibohra/TuneKit

Upload your data → Get a fine-tuned SLM. Free.

Category
fine tuning
Stars
150
Readiness
ready (81/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning llm llms machine-learning open-source slms

Saivineeth147/lora-speedrun

Speedrunning LoRA fine-tuning: frozen task, frozen hardware, public wall-clock leaderboard. modded-nanogpt for fine-tuning.

Category
fine tuning
Stars
144
Readiness
ready (88/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

benchmark fine-tuning leaderboard llm lora peft

kubeflow/sdk

Universal Python SDK to run AI workloads on Kubernetes

Category
fine tuning
Stars
136
Readiness
ready (81/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +5 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai distributed fine-tuning huggingface hyperparameter-optimization jax

kennethleungty/Finance-LLMs

Comprehensive Compilation of Real-World LLM & AI Agent Use Cases in Financial Services

Category
fine tuning
Stars
135
Readiness
ready (93/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent agent-skill agentic agentic-ai agentic-workflow agents

arkorlab/arkor

The Best TypeScript framework for fine-tuning

Category
fine tuning
Stars
131
Readiness
ready (89/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +79 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai developer-tools fine-tuning gpu javascript llm

llmsresearch/llm-flashcards

Visual knowledge bank for understanding large language models, with 180 concept cards from tokenization to deployment.

Category
fine tuning
Stars
126
Readiness
ready (80/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +13 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agents ai anki attention deep-learning fine-tuning

NVlabs/alpamayo-recipes

Developer Hub for NVIDIA Alpamayo, containing ready-to-use recipes for fine-tuning, reinforcement-learning post-training, quantization, and deployment.

Category
fine tuning
Stars
106
Readiness
ready (96/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +23 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

alpamayo autonomous-driving autonomous-vehicles fine-tuning nvidia physical-ai

ldbc/Text2GraphQuery-DataGen

Fine-Tuning Dataset Auto-Generation for Graph Query Languages.

Category
fine tuning
Stars
102
Readiness
ready (87/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

awesome-list fine-tuning graphdb hacktoberfest llm peft

multimindlab/multimind-sdk

Your SDK solves all of this. One interface. Unified logic. Local + hosted models. Fine-tuning. Agent tools. Enterprise-ready. Hybrid RAG.Star 🌟 if you like it!

Category
fine tuning
Stars
93
Readiness
needs review (71/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agentic-ai agentic-framework agents ai ai-agents deep-learning

Red-Hat-AI-Innovation-Team/training_hub

An algorithm-focused interface for common llm training, continual learning, and reinforcement learning techniques

Category
fine tuning
Stars
88
Readiness
ready (95/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-agents fine-tuning language-model optimization training

Edoar-do/HuBERT-ECG

A self-supervised foundation ECG model for broad and scalable cardiac applications

Category
fine tuning
Stars
80
Readiness
ready (78/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

artificial-inte deep-learning ecg fine-tuning foundation-models self-supervised-learning

DaoyuanLi2816/llama3-emotion-lora

Emotion text classification using Llama3-8b with LoRA and FlashAttention. Based on LLaMA-Factory.

Category
fine tuning
Stars
73
Readiness
ready (77/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

emotion-classification fine-tuning flash-attention llama-factory llama3 llm

speediedan/finetuning-scheduler

A PyTorch Lightning extension that accelerates and enhances foundation model experimentation with flexible fine-tuning schedules.

Category
fine tuning
Stars
70
Readiness
ready (88/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

artificial-intelligence fine-tuning finetuning machine-learning neural-networks pytorch

peremartra/Rearchitecting-LLMs

Official code for the Manning book on structural LLM optimization: depth/width pruning, knowledge distillation, and attention optimization, runnable on free Colab GPUs.

Category
fine tuning
Stars
66
Readiness
ready (97/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +7 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-fairness attention-optimization fine-tuning knowledge-distilation large-language-models llm

SauravP97/ai-engineering-primer

Learn Agentic AI, Deep Learning, Agentic Engineering, RAG and Multi-Agent Workflows.

Category
fine tuning
Stars
60
Readiness
ready (95/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agentic-ai agentic-workflow ai-engineering course fine-tuning google

lightning-rod-labs/lightningrod-python-sdk

Better AI Predictions ⚡

Category
fine tuning
Stars
57
Readiness
ready (82/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

dataset-generation fine-tuning llm-datasets llm-tooling synthetic-data

Mattral/production-vlm-engineering

Reproducible, production-grade pipelines for modern multimodal vision systems. Efficient VLM adaptation · Embedding-space drift detection · Edge inference · Robustness & safety

Category
fine tuning
Stars
51
Readiness
ready (81/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

edge-inference fine-tuning vision-language-model

FredyRivera-dev/claude_converter

A zero-dependency Python module for inspecting and converting coding-agent session files (.jsonl) — Claude Code, Codex, and Pi — into the messages format expected by Hugging Face Transformers.

Category
fine tuning
Stars
46
Readiness
ready (74/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

anthropic claude claude-code codex codex-cli converter

QingGo/engram-peft

🚀 Engram-PEFT: An unofficial implementation of DeepSeek Engram. Inject high-capacity conditional memory into LLMs via sparse retrieval PEFT without increasing inference FLOPs / DeepSeek Engram 架构的非官方实

Category
fine tuning
Stars
41
Readiness
ready (84/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

deepseek engram fine-tuning llm peft pytorch

AcTePuKc/Universal-TTS-Guide

This comprehensive guide provides a universal process for preparing your own speech datasets and training a custom Text-to-Speech (TTS) model.

Category
fine tuning
Stars
37
Readiness
needs review (66/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

audio-processing dataset-preparation documentation fine-tuning guide how-to

Koratahiu/Advanced_Optimizers

A family of highly efficient, lightweight yet powerful optimizers.

Category
fine tuning
Stars
36
Readiness
ready (90/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

adam-optimizer compression fine-tuning llm low-rank-approximation memory-efficient

DaoyuanLi2816/llm-gpu-lab

One GPU. Full LLM workflow. Real benchmarks. No cloud required.

Category
fine tuning
Stars
36
Readiness
ready (79/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning gguf gpu llama-cpp llm lora

anote-ai/Panacea

Panacea is a framework for building collaborative, intelligent multi agent AI systems. The framework provides a robust infrastructure for creating and managing multiple AI agents, and enables develope

Category
fine tuning
Stars
33
Readiness
ready (81/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agents ai automation core-product deep-learning fine-tuning

Hysocs/Aozora_Trainer

A highly memory-efficient fine-tuning toolkit for SDXL and ANIMA, combining modern VRAM-saving techniques with custom optimizers to enable full-quality model training on GPUs with as little as 12 GB o

Category
fine tuning
Stars
32
Readiness
ready (84/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

12gb-vram ai ai-fine-tuning ai-finetune ai-training anima

DaoyuanLi2816/tracedistill

Distill teacher chains-of-thought into a LoRA adapter via a strict boxed-answer format contract + two-phase Train→Nudge (silver-medal NVIDIA Nemotron reasoning recipe, as a tested library).

Category
fine tuning
Stars
32
Readiness
ready (86/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chain-of-thought fine-tuning kaggle llm lora mamba

jcwang587/cgcnn2

Reproduction of CGCNN for predicting material properties

Category
fine tuning
Stars
28
Readiness
needs review (71/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

computational-materials fine-tuning graph-model python

i4Ds/whisper-finetune

This repository contains code for fine-tuning the Whisper speech-to-text model.

Category
fine tuning
Stars
24
Readiness
ready (95/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning nlp speech-to-text whisper

Aisuko/notebooks

Implementation for the different ML tasks on Kaggle platform with GPUs.

Category
fine tuning
Stars
24
Readiness
ready (91/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

accelerator computer-vision fine-tuning kaggle large-language-models multimodal

gvkhosla/pi-tinker

Fine-tune open-source models with Tinker from inside Pi — managed improve loops, data prep, evals, smoke tests, deploy snippets, and checkpoint chat.

Category
fine tuning
Stars
24
Readiness
ready (78/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning llm open-source-models pi-package post-training sft

Agnuxo1/openclaw-seed

OpenCLAW SEED — Autonomous self-evolving research agent with LoRA fine-tuning. Part of P2PCLAW ecosystem. Trains itself on research papers, improves over time. Apache 2.0.

Category
fine tuning
Stars
23
Readiness
ready (88/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-agents autonomous-agents fine-tuning huggingface llm lora

cloneofsimo/lora

Using Low-rank adaptation to quickly fine-tune diffusion models.

Category
fine tuning
Stars
7,545
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 868 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

diffusion dreambooth fine-tuning lora stable-diffusion

LLMBook-zh/LLMBook-zh.github.io

《大语言模型》作者：赵鑫，李军毅，周昆，唐天一，文继荣

Category
fine tuning
Stars
4,536
Readiness
high risk (41/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
19/100 · low confidence

Why: +17 stars in 30 days

Why it may be a gem: limited evidence; inspect maintenance signals before adopting

Strongest signals: issue load, documentation. Risks: no push in 339 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

artificial-intelligence deep-learning deep-neural-networks deep-reinforcement-learning fine-tuning language-model

hiyouga/ChatGLM-Efficient-Tuning

Fine-tuning ChatGLM-6B with PEFT | 基于 PEFT 的高效 ChatGLM 微调

Category
fine tuning
Stars
3,717
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: repository is archived, no push in 1030 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

alpaca chatglm chatglm2 chatgpt fine-tuning huggingface

ZhaoJ9014/face.evoLVe

🔥🔥High-Performance Face Recognition Library on PaddlePaddle & PyTorch🔥🔥

Category
fine tuning
Stars
3,587
Readiness
needs review (60/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 505 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

artificial-intelligence computer-vision convolutional-neural-network data-augmentation deep-learning face-alignment

sentient-agi/OML-1.0-Fingerprinting

OML 1.0 via Fingerprinting: Open, Monetizable, and Loyal AI

Category
fine tuning
Stars
3,502
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 561 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning fingerprint loyalty oml sentient verifiable-ai

dbiir/UER-py

Open Source Pre-training Model Framework in PyTorch & Pre-trained Model Zoo

Category
fine tuning
Stars
3,111
Readiness
needs review (57/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 820 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

albert bart bert chinese classification clue

ashishpatel26/LLM-Finetuning

LLM Finetuning with peft

Category
fine tuning
Stars
2,973
Readiness
needs review (50/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: limited evidence; inspect maintenance signals before adopting

Strongest signals: issue load, fork interest, documentation. Risks: no push in 371 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

falcon fine-tuning huggingface llama llama2 llm

stochasticai/xTuring

Build, personalize and control your own LLMs. From data pre-processing to fine-tuning, xTuring provides an easy way to personalize open-source LLMs. Join our discord community: https://discord.gg/TgHX

Category
fine tuning
Stars
2,673
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
26/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 156 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

adapter deep-learning fine-tuning finetuning gen-ai generative-ai

adithya-s-k/AI-Engineering.academy

Mastering Applied AI, One Concept at a Time

Category
fine tuning
Stars
2,370
Readiness
needs review (55/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
27/100 · low confidence

Why: +5 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 161 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning finetuning finetuning-llms inference large-language-models llm

YiVal/YiVal

Your Automatic Prompt Engineering Assistant for GenAI Applications

Category
fine tuning
Stars
2,133
Readiness
needs review (57/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 837 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai ai-experiments ai-toolkit aigc api auto-prompting

eosphoros-ai/DB-GPT-Hub

A repository that contains models, datasets, and fine-tuning techniques for DB-GPT, with the purpose of enhancing model performance in Text-to-SQL

Category
fine tuning
Stars
2,003
Readiness
needs review (55/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +6 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 401 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

database datasets fine-tuning gpt hacktoberfest llm

adobe-research/custom-diffusion

Custom Diffusion: Multi-Concept Customization of Text-to-Image Diffusion (CVPR 2023)

Category
fine tuning
Stars
1,977
Readiness
high risk (44/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

computer-vision customization diffusion-models few-shot fine-tuning pytorch

google-deepmind/penzai

A JAX research toolkit for building, editing, and visualizing neural networks.

Category
fine tuning
Stars
1,898
Readiness
needs review (51/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 411 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning interpretability jax neural-networks visualization

THUDM/LongWriter

[ICLR 2025] LongWriter: Unleashing 10,000+ Word Generation from Long Context LLMs

Category
fine tuning
Stars
1,871
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 409 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning llm long-context long-text

ray-project/llm-applications

A comprehensive guide to building RAG-based LLM applications for production.

Category
fine tuning
Stars
1,858
Readiness
needs review (48/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 736 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

anyscale fine-tuning llama2 llms machine-learning openai

tianrun-chen/SAM-Adapter-PyTorch

Adapting Meta AI's Segment Anything to Downstream Tasks with Adapters and Prompts

Category
fine tuning
Stars
1,545
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +3 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

2d-segmentation adapter camouflage-images camouflaged-object-detection camouflaged-target-detection fine-tune

jina-ai/finetuner

:dart: Task-oriented embedding tuning for BERT, CLIP, etc.

Category
fine tuning
Stars
1,505
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: repository is archived, no push in 879 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bert few-shot-learning fine-tuning finetuning jina metric-learning

SakanaAI/text-to-lora

Hypernetworks that adapt LLMs for specific benchmark tasks using only textual task description as the input

Category
fine tuning
Stars
1,298
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 425 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning hypernetworks llm lora machine-learning

uclaml/SPIN

The official implementation of Self-Play Fine-Tuning (SPIN)

Category
fine tuning
Stars
1,250
Readiness
needs review (51/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 821 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

deep-learning fine-tuning large-language-models self-play

AGI-Edgerunners/LLM-Adapters

Code for our EMNLP 2023 Paper: "LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of Large Language Models"

Category
fine tuning
Stars
1,235
Readiness
needs review (51/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 880 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

adapters fine-tuning large-language-models parameter-efficient

datadreamer-dev/DataDreamer

DataDreamer: Prompt. Generate Synthetic Data. Train & Align Models.   🤖💤

Category
fine tuning
Stars
1,115
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 551 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

alignment deep-learning fine-tuning gpt instruction-tuning llm

Tencent/TencentPretrain

Tencent Pre-training framework in PyTorch & Pre-trained Model Zoo

Category
fine tuning
Stars
1,090
Readiness
needs review (47/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 733 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

albert bart bert chinese classification clue

RL-VIG/LibFewShot

[TPAMI 2023] LibFewShot: A Comprehensive Library for Few-shot Learning.

Category
fine tuning
Stars
1,071
Readiness
needs review (59/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 284 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

few-shot-learning fine-tuning image-classification meta-learning pytorch

datawhalechina/base-llm

从 NLP 到 LLM 的算法全栈教程，在线阅读地址：https://datawhalechina.github.io/base-llm/

Category
fine tuning
Stars
953
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +20 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bert deeplearning docker fine-tuning linux llama

lonePatient/Bert-Multi-Label-Text-Classification

This repo contains a PyTorch implementation of a pretrained BERT model for multi-label text classification.

Category
fine tuning
Stars
921
Readiness
needs review (60/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 1207 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

albert bert fine-tuning multi-label-classification nlp pytorch

AutoArk/TinyEngram

Research of DeepSeek Engram Architecture based on Qwen-3 and Stable Diffusion series.

Category
fine tuning
Stars
899
Readiness
high risk (40/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +124 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

deepseek deepseek-ai engram fine-tuning llm llm-memory

georgian-io/LLM-Finetuning-Toolkit

Toolkit for fine-tuning, ablating and unit-testing open-source LLMs.

Category
fine tuning
Stars
870
Readiness
needs review (55/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
16/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 95 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ablation-study classification falcon fine-tuning finetuning flan-t5

theredsix/cerebellum

Browser automation system that uses AI-driven planning to navigate web pages and perform goals.

Category
fine tuning
Stars
867
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
70/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: repository is archived. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agents ai automation-framework automation-ui browser fine-tuning

dvgodoy/FineTuningLLMs

Official repository of my book "A Hands-On Guide to Fine-Tuning LLMs with PyTorch and Hugging Face"

Category
fine tuning
Stars
850
Readiness
needs review (56/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
0/100 · low confidence

Why: +5 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bitsandbytes fine-tuning finetuning finetuning-llms hugging-face huggingface

haofanwang/Lora-for-Diffusers

The most easy-to-understand tutorial for using LoRA (Low-Rank Adaptation) within diffusers framework for AI Generation Researchers🔥

Category
fine tuning
Stars
824
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 849 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

aigc colossalai diffusers fine-tuning guidebook lora

MrGiovanni/ModelsGenesis

[MICCAI 2019 Young Scientist Award] [MEDIA 2020 Best Paper Award] Models Genesis, one of the first "foundation" models in medical image analysis for multiple downstream tasks

Category
fine tuning
Stars
789
Readiness
needs review (49/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +3 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 411 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

3d-model fine-tuning foundation-models pre-trained-model representation-learning self-supervised-learning

AGI-Arena/MARS

The official implementation of MARS: Unleashing the Power of Variance Reduction for Training Large Models

Category
fine tuning
Stars
724
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
22/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 134 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning large-language-models optimization-algorithms optimizer pretraining

Coobiw/MPP-LLaVA

Personal Project: MPP-Qwen14B & MPP-Qwen-Next(Multimodal Pipeline Parallel based on Qwen-LM). Support [video/image/multi-image] {sft/conversations}. Don't let the poverty limit your imagination! Train

Category
fine tuning
Stars
685
Readiness
high risk (39/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 515 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

deepspeed fine-tuning mllm model-parallel multimodal-large-language-models pipeline-parallelism

uclaml/SPPO

The official implementation of Self-Play Preference Optimization (SPPO)

Category
fine tuning
Stars
590
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 561 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

deep-learning fine-tuning large-language-models rlhf self-play

wpydcr/LLM-Kit

🚀WebUI integrated platform for latest LLMs | 各大语言模型的全流程工具 WebUI 整合包。支持主流大模型API接口和开源模型。支持知识库，数据库，角色扮演，mj文生图，LoRA和全参数微调，数据集制作，live2d等全流程应用工具

Category
fine tuning
Stars
553
Readiness
needs review (46/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 255 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chatbot embeddings fine-tuning generative-agents llm player

Curated-Awesome-Lists/awesome-llms-fine-tuning

Explore a comprehensive collection of resources, tutorials, papers, tools, and best practices for fine-tuning Large Language Models (LLMs). Perfect for ML practitioners and researchers!

Category
fine tuning
Stars
525
Readiness
high risk (44/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
25/100 · low confidence

Why: +3 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 613 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai awesome-list deep-learning fine-tuning gpt large-language-models

THUDM/LongCite

LongCite: Enabling LLMs to Generate Fine-grained Citations in Long-context QA

Category
fine tuning
Stars
519
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 584 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

benchmark citation-generation fine-tuning llm long-context

Leeroo-AI/mergoo

A library for easily merging multiple LLM experts, and efficiently train the merged LLM.

Category
fine tuning
Stars
516
Readiness
high risk (44/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 711 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

artificial-intelligence fine-tuning generative-ai large-language-models llm lora

zhao-kun/VibeVoiceFusion

VibeVoiceFusion is a full-stack, multi-speaker voice generation web system featuring LoRA fine-tuning, batch generation, and VRAM optimization. Based on Microsoft's VibeVoice (AR + diffusion architect

Category
fine tuning
Stars
488
Readiness
high risk (43/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
27/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 165 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

aigc autoregressive-models fine-tuning language-model lora speech-synthesis

Joyce94/LLM-RLHF-Tuning

LLM Tuning with PEFT (SFT+RM+PPO+DPO with LoRA)

Category
fine tuning
Stars
453
Readiness
high risk (39/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 1031 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning language-model llama llm lora peft

okuvshynov/slowllama

Finetune llama2-70b and codellama on MacBook Air without quantization

Category
fine tuning
Stars
449
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: repository is archived, no push in 862 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

apple-silicon fine-tuning llama llama2

MuLabPKU/PiSSA

PiSSA: Principal Singular Values and Singular Vectors Adaptation of Large Language Models(NeurIPS 2024 Spotlight)

Category
fine tuning
Stars
429
Readiness
high risk (34/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load. Risks: no push in 403 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning peft quantization

tigerlab-ai/tiger

Open Source LLM toolkit to build trustworthy LLM applications. TigerArmor (AI safety), TigerRAG (embedding, RAG), TigerTune (fine-tuning)

Category
fine tuning
Stars
404
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 979 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-safety aisafety classification data-augmentation fine-tuning large-language-models

Shivanandroy/simpleT5

simpleT5 is built on top of PyTorch-lightning⚡️ and Transformers🤗 that lets you quickly train your T5 models.

Category
fine tuning
Stars
403
Readiness
needs review (55/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 1176 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

classification fine-tuning finetune pytorch simplet5 summarization

zju-vipa/Odyssey

Odyssey: Empowering Minecraft Agents with Open-World Skills

Category
fine tuning
Stars
399
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +3 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 289 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent embodied-agent fine-tuning large-language-model large-language-models llm

HIT-SCIR/huozi

活字通用大模型

Category
fine tuning
Stars
395
Readiness
needs review (50/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 694 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning large-language-models llm nlp

bentoml/BentoDiffusion

BentoDiffusion: A collection of diffusion models served with BentoML

Category
fine tuning
Stars
388
Readiness
ready (75/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai diffusion-models fine-tuning kubernetes lora model-serving

ml6team/fondant

Production-ready data processing made easy and shareable

Category
fine tuning
Stars
358
Readiness
needs review (50/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
28/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 168 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

data-processing fine-tuning foundation-models machine-learning pipeline python

ddzipp/AutoAudit

AutoAudit—— the LLM for Cyber Security 网络安全大语言模型

Category
fine tuning
Stars
356
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 525 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

cyber-security fine-tuning gpt llama lora qlora

ReinFlow/ReinFlow

[NeurIPS 2025] Flow x RL. "ReinFlow: Fine-tuning Flow Policy with Online Reinforcement Learning". Support VLAs e.g., Pi0, Pi0.5, GR00TN1.5. Fully open-sourced.

Category
fine tuning
Stars
355
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
17/100 · low confidence

Why: +4 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 105 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

actorcritic fine-tuning finetuning-rl finetuning-vision-models flow flowmatching

WangRongsheng/MedQA-ChatGLM

🛰️ 基于真实医疗对话数据在ChatGLM上进行LoRA、P-Tuning V2、Freeze、RLHF等微调，我们的眼光不止于医疗问答

Category
fine tuning
Stars
340
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: repository is archived, no push in 1070 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chatglm-6b chatgpt dataset fine-tuning freeze huggingface

Zeyi-Lin/Qwen3-Medical-SFT

Qwen3 Fine-tuning: Medical R1 Style Chat

Category
fine tuning
Stars
333
Readiness
high risk (43/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 434 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning qwen3 r1 sft

ywyue/FiT3D

[ECCV 2024] Improving 2D Feature Representations by 3D-Aware Fine-Tuning

Category
fine tuning
Stars
329
Readiness
needs review (51/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 229 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

3d-awareness clip deit-iii dinov2 fine-tuning foundation-models

Epistates/pmetal

PMetal: high-performance Apple Silicon framework for local LLM inference, LoRA/QLoRA fine-tuning, serving, quantization, and MLX/Metal acceleration.

Category
fine tuning
Stars
306
Readiness
high risk (44/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +4 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai ane apple-silicon deep-learning distillation fine-tuning

anchen1011/FireAct

FireAct: Toward Language Agent Fine-tuning

Category
fine tuning
Stars
296
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 1020 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent fine-tuning large-language-models llm reasoning

cliang1453/BOND

BOND: BERT-Assisted Open-Domain Name Entity Recognition with Distant Supervision

Category
fine tuning
Stars
290
Readiness
needs review (55/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 1892 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bert dataset distant-supervision fine-tuning named-entity-recognition natural-language-processing

MaryamBoneh/Vehicle-Detection

Vehicle Detection Using Deep Learning and YOLO Algorithm

Category
fine tuning
Stars
286
Readiness
needs review (51/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 1050 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

car-counting dataset deep-learning fine-tuning image-processing object-detection

mrm8488/shared_colab_notebooks

A Repo to store the Google Colaboratory Notebooks that I have created and shared

Category
fine tuning
Stars
284
Readiness
needs review (49/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
28/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 499 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai artificial-neural-networks cv fine-tuning gans google-colaboratory-notebooks

ystemsrx/Qwen2-Boundless

A fine-tuned model from Qwen2-1.5B-Instruct, capable of handling sensitive topics like violence, explicit content. / 从 Qwen2-1.5B-Instruct 微调，能处理各类敏感话题

Category
fine tuning
Stars
283
Readiness
needs review (58/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 715 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent ai fine-tuning qwen2 sensitive text2text

zer0int/CLIP-fine-tune

Fine-tuning code for CLIP models

Category
fine tuning
Stars
275
Readiness
high risk (41/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: limited evidence; inspect maintenance signals before adopting

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

clip comfyui fine-tune fine-tuning finetune openai

anakin87/qwen-scheduler-grpo

Train a Language Model with GRPO to create a schedule from a list of events and priorities

Category
fine tuning
Stars
272
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
0/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning grpo llm post-training reasoning reinforcement-learning

TYH-labs/unsloth-buddy

Zero-friction LLM fine-tuning skill for Claude Code, Gemini CLI & any ACP agent. Unsloth on NVIDIA · TRL+MPS/MLX on Apple Silicon. Automates env setup, LoRA training (SFT, DPO, GRPO, vision), post-hoc

Category
fine tuning
Stars
270
Readiness
needs review (64/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

apple-silicon claude-code dpo fine-tuning gaslamp grpo

sakuranew/BERT-AttributeExtraction

USING BERT FOR Attribute Extraction in KnowledgeGraph. fine-tuning and feature extraction. 使用基于bert的微调和特征提取方法来进行知识图谱百度百科人物词条属性抽取。

Category
fine tuning
Stars
266
Readiness
needs review (50/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 2685 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai attribute-extraction bert deeplearning feature-extraction fine-tuning

iamarunbrahma/finetuned-qlora-falcon7b-medical

Finetuning of Falcon-7B LLM using QLoRA on Mental Health Conversational Dataset

Category
fine tuning
Stars
263
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 935 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chatbot chatbots conversational-ai falcon falcon-7b fine-tuning

JJLibra/CC-Pan

🤗 Official implementation for "CC-Pan: Channel-wise Compression based Diffusion for Efficient Pan-Sharpening" https://arxiv.org/abs/2602.04473

Category
fine tuning
Stars
261
Readiness
needs review (55/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning image-fusion pansharpening remote-sensing stable-diffusion

WangRongsheng/Aurora

The official codes for "Aurora: Activating chinese chat capability for Mixtral-8x7B sparse Mixture-of-Experts through Instruction-Tuning"

Category
fine tuning
Stars
260
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 820 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chinese fine-tuning gpt instruction-tuning language-model large-language-models

ystemsrx/Qwen2.5-Sex

A fine-tuned model from Qwen2.5-1.5B-Instruct, capable of handling sensitive topics. / 从 Qwen2.5-1.5B-Instruct 微调，主要擅长处理色情话题

Category
fine tuning
Stars
253
Readiness
needs review (57/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 670 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai erotic fine-tuning porn sensitive text2text

promptslab/LLMtuner

FineTune LLMs in few lines of code (Text2Text, Text2Speech, Speech2Text)

Category
fine tuning
Stars
250
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 937 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning fine-tuning-llm finetune finetune-gpt finetune-llama finetune-llm

git-disl/awesome_LLM-harmful-fine-tuning-papers

A survey on harmful fine-tuning attack for large language model (ACM CSUR)

Category
fine tuning
Stars
247
Readiness
needs review (47/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

alignment attack defense emergent fine-tuning finetuning

backprop-ai/backprop

Backprop makes it simple to use, finetune, and deploy state-of-the-art ML models.

Category
fine tuning
Stars
240
Readiness
high risk (43/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 1922 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bert fine-tuning image-classification language-model multilingual-models natural-language-processing

AoqunJin/Awesome-VLA-Post-Training

A collection of vision-language-action model post-training methods.

Category
fine tuning
Stars
235
Readiness
needs review (51/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

embodied-agent embodied-ai fine-tuning post-training vision-language-action-model vla

WangRongsheng/SAM-fine-tune

🌌 Fine tune specific SAM model on any task

Category
fine tuning
Stars
228
Readiness
high risk (39/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 676 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning sam segment-anything segmentation

gyunggyung/KoGPT2-FineTuning

🔥 Korean GPT-2, KoGPT2 FineTuning cased. 한국어 가사 데이터 학습 🔥

Category
fine tuning
Stars
228
Readiness
needs review (62/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 465 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning finetuning gpt2 kogpt2 korean korean-nlp

myCSAI/Ziwei

基于已有基座模型微调的算命大模型

Category
fine tuning
Stars
224
Readiness
high risk (44/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +4 stars in 30 days

Why it may be a gem: strong signals despite limited visibility

Strongest signals: issue load, fork interest. Risks: no push in 955 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai aigc fine-tuning

codelion/ellora

Enhancing LLMs with LoRA

Category
fine tuning
Stars
224
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, documentation, license. Risks: no push in 291 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

accuracy-analysis chain-of-thought chain-of-thought-reasoning data-generation distillation fine-tune

YuanchenBei/Awesome-Pretraining-for-Graph-Neural-Networks

A curated list of papers on pre-training for graph neural networks (Pre-train4GNN).

Category
fine tuning
Stars
215
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
24/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, documentation, license. Risks: no push in 584 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning graph graph-algorithms graph-neural-networks graph-pretraining pre-trained-model

zeyadusf/LLMs-from-Scratch

Build a Large Language Model (From Scratch) book and Finetuned Models

Category
fine tuning
Stars
209
Readiness
needs review (50/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, fork interest, documentation. Risks: no push in 681 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

advanced-neural-network attention-mechanism bert book classification coding

hscspring/hcgf

Humanable Chat Generative-model Fine-tuning | LLM微调

Category
fine tuning
Stars
205
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, documentation, license. Risks: no push in 1050 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chatglm chatglm2 chatgpt fine-tuning large-language-models llama

yueyu1030/COSINE

[NAACL 2021] This is the code for our paper `Fine-Tuning Pre-trained Language Model with Weak Supervision: A Contrastive-Regularized Self-Training Approach'.

Category
fine tuning
Stars
205
Readiness
needs review (56/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, documentation, license. Risks: no push in 1451 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agnews contrastive-learning dataset fine-tuning language-model learning-with-noisy-labels

kamalkraj/ALBERT-TF2.0

ALBERT model Pretraining and Fine Tuning using TF2.0

Category
fine tuning
Stars
204
Readiness
needs review (58/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, fork interest, documentation. Risks: no push in 1232 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

albert albert-tf2 classifier cola fine-tuning glue

charlesdedampierre/BunkaTopics

🗺️ Data Cleaning and Textual Data Visualization 🗺️

Category
fine tuning
Stars
201
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, documentation, license. Risks: no push in 441 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

cartography data-cleaning explainability fine-tuning llms machine-learning

LAION-AI/scaling-laws-openclip

Reproducible scaling laws for contrastive language-image learning (https://arxiv.org/abs/2212.07143)

Category
fine tuning
Stars
201
Readiness
high risk (40/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, documentation. Risks: no push in 412 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

clip deep-learning few-shot-learning fine-tuning laion openclip

modelscope/ms-swift

Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 600+ LLMs (Qwen3.6, DeepSeek-V4, GLM-5.1, InternLM3, Llama4, ...) and 300+ MLLMs (Qwen3-VL, Qwen3-Omni, InternVL3.5, Ovis2.5, GLM4.5v, Gemma4, Llava, Phi

Category
fine tuning
Stars
15,083
Readiness
ready (92/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +66 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

deepseek-r1 embedding grpo internvl liger llama

OpenPipe/ART

Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen3.6, GPT-OSS, Llama, and more!

Category
fine tuning
Stars
10,563
Readiness
ready (91/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +6 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent agentic-ai grpo llms lora qwen

meshtastic/firmware

The official firmware for Meshtastic, an open-source, off-grid mesh communication system.

Category
fine tuning
Stars
8,079
Readiness
ready (90/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +46 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

esp32 gps heltec hiking lora mesh

markqvist/Reticulum

The cryptography-based networking stack for building unstoppable networks with LoRa, Packet Radio, WiFi and everything in between.

Category
fine tuning
Stars
6,538
Readiness
ready (75/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +65 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora mesh mesh-networks network networking-stack packet-radio

ExpressLRS/ExpressLRS

High Performance Open Source Radio Control Link

Category
fine tuning
Stars
5,218
Readiness
ready (89/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +30 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

drone esp32 esp8266 fpv lora lr1121

1technophile/OpenMQTTGateway

MQTT gateway for ESP8266 or ESP32 with bidirectional 433mhz/315mhz/868mhz, Infrared communications, BLE, Bluetooth, beacons detection, mi flora, mi jia, LYWSD02, LYWSD03MMC, Mi Scale, TPMS, BBQ thermo

Category
fine tuning
Stars
4,073
Readiness
ready (87/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +4 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

arduino arduino-uno ble ble-beacons bridge esp32

absmach/magistrala

IoT Platform Framework

Category
fine tuning
Stars
2,623
Readiness
ready (98/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

coap distributed-systems edge edge-computing golang iiot

cyberman54/ESP32-Paxcounter

Wifi & BLE driven passenger flow metering with cheap ESP32 boards

Category
fine tuning
Stars
2,074
Readiness
ready (90/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

arduino bme680 dcf77 esp32 gps gps-tracker

dphnAI/sonar

Large-scale LLM inference engine

Category
fine tuning
Stars
1,823
Readiness
ready (81/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +7 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

api-rest cuda inference-engine inferentia intel lora

meshtastic/Meshtastic-Android

Android application for Meshtastic

Category
fine tuning
Stars
1,775
Readiness
ready (90/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +15 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

android bluetooth-low-energy encrypted-messaging esp32 gps iot

markqvist/Sideband

LXMF client for Android, Linux and macOS allowing you to communicate with people or LXMF-compatible systems over Reticulum networks using LoRa, Packet Radio, WiFi, I2P, or anything else Reticulum supp

Category
fine tuning
Stars
1,665
Readiness
ready (75/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +21 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

android lora lxmf lxst mesh p2p

TheThingsNetwork/lorawan-stack

The Things Stack, an Open Source LoRaWAN Network Server

Category
fine tuning
Stars
1,157
Readiness
ready (94/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

iot lora lorawan ttn

chirpstack/chirpstack

ChirpStack open-source LoRaWAN Network Server

Category
fine tuning
Stars
1,085
Readiness
ready (97/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chirpstack internet-of-things iot lora lorawan

lyusupov/SoftRF

:airplane: Multi-functional, compatible DIY general aviation proximity awareness system

Category
fine tuning
Stars
997
Readiness
ready (92/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ads-b aprs aviation esp32 flarm gliding

torlando-tech/columba

Native Android messaging app using Bluetooth LE, TCP, or RNode (LoRa) over LXMF and Reticulum

Category
fine tuning
Stars
956
Readiness
ready (81/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +13 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ble lora lxmf mesh-networks off-grid privacy

Finrandojin/alexandria-audiobook

AI-powered multi-voice audiobook generator — LLM script annotation, voice cloning, voice design, LoRA training, per-line style control, and export to MP3, chaptered M4B, or Audacity multi-track. Built

Category
fine tuning
Stars
857
Readiness
ready (90/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +9 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai audiobook audiobook-generator audiobookshelf chapter-markers dialogue-generation

sunrainyg/RandOpt

Official Codebase for "Neural Thickets: Diverse Task Experts Are Dense Around Pretrained Weights" (ICML 2026 Spotlight)

Category
fine tuning
Stars
633
Readiness
needs review (63/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

black-box-optimization ensemble-learning large-language-models llm lora neuroevolution

wildminder/awesome-ltx2

All available LTX-2 models, encoders, workflows, LoRAs for ComfyUI

Category
fine tuning
Stars
573
Readiness
ready (74/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +9 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai comfyui itv lora ltx-2 ltxv

lukeswitz/AntiHunter

AntiHunter Perimeter Defense Systems - DIGI Node Firmware

Category
fine tuning
Stars
560
Readiness
ready (85/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +5 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

detection esp32 lora mesh security sigint

richonguzman/LoRa_APRS_iGate

LoRa APRS iGATE for ESP32 Based Board with Rx + Tx capabilities

Category
fine tuning
Stars
509
Readiness
ready (73/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +6 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

aprs aprs-igate aprs-is aprs-tracker arduino digipeater

richonguzman/LoRa_APRS_Tracker

LoRa APRS Tracker with Tx and Rx capabilities, Messages, Wx, Winlink and more...

Category
fine tuning
Stars
502
Readiness
ready (73/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

aprs aprs-igate aprs-is aprs-tracker aprsis arduino

lora-rs/lora-rs

LoRa and LoRaWAN crates for End Devices

Category
fine tuning
Stars
464
Readiness
ready (97/100 heuristic points; not a probability)
Data confidence
medium
Maintainer health
healthy
Maintenance risk
0/100 · high confidence

Why: +1 stars in 7 days; 34 commits in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; consistent human and community activity; open issue backlog is stable or shrinking

Strongest signals: push recency, commit activity, contributor breadth. Risks: None identified. Missing inputs: release recency.

embedded-rust embedded-systems iot iot-device lora lorawan

chirpstack/chirpstack-gateway-bridge

ChirpStack Gateway Bridge abstracts Packet Forwarder protocols into Protobuf or JSON over MQTT.

Category
fine tuning
Stars
445
Readiness
needs review (63/100 heuristic points; not a probability)
Data confidence
high
Maintainer health
healthy
Maintenance risk
8/100 · high confidence

Why: +2 stars in 30 days; 34 lifetime contributors

Why it may be a gem: healthy maintenance and project fundamentals; open issue backlog is stable or shrinking

Strongest signals: contributor breadth, issue load, fork interest. Risks: no pull-request review responses in 30 days. Missing inputs: None.

chirpstack json lora lorawan mqtt protobuf

ClusterDuck-Protocol/ClusterDuck-Protocol

Firmware for an ad-hoc mesh network of Internet-of-Things devices based on LoRa (Long Range radio) that can be deployed quickly and at low cost.

Category
fine tuning
Stars
445
Readiness
needs review (67/100 heuristic points; not a probability)
Data confidence
high
Maintainer health
watch
Maintenance risk
34/100 · high confidence

Why: +2 stars in 7 days; 25 lifetime contributors

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, contributor breadth, issue load. Risks: recent commit cadence is 0/2.0 of its monthly baseline, no pull-request review responses in 30 days, no maintainer response activity in 30 days. Missing inputs: None.

call-for-code ducks emergency-network hacktoberfest iot lora

baidu-baige/LoongForge

A unified, high-performance framework for training LLMs, VLMs, diffusion, and embodied models on NVIDIA GPUs and Kunlun XPUs.

Category
fine tuning
Stars
339
Readiness
ready (91/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +12 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

diffusion distributed gpu kunlun llm lora

LuqP2/Image-MetaHub

A desktop application for browsing, searching, and organizing AI-generated images locally. Designed for performance with large collections, focusing on powerful metadata filtering and complete privacy

Category
fine tuning
Stars
310
Readiness
ready (80/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-art ai-tools automatic1111 comfyui electron flux

rlops/rlix

Run more RL experiments. Wait less for GPUs.

Category
fine tuning
Stars
294
Readiness
ready (86/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agentic-rl gpu-scheduling llm-training lora ml-systems mlops

LoRaMesher/LoRaMesher

The LoRaMesher library implements a distance-vector routing protocol for communicating messages among LoRa nodes.

Category
fine tuning
Stars
281
Readiness
ready (93/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

arduino esp-idf esp32 iot-service-network lora mesh

RAKWireless/WisBlock

Quickstart, tutorials and examples for the RAKwireless WisBlock product line.

Category
fine tuning
Stars
245
Readiness
ready (90/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

esp32-s3 esp32-wrover lora lorawan nrf52 rp2040

openhop-dev/openhop_repeater

A lightweight Repeater Daemon implemented in Python, built using the openhop_core library.

Category
fine tuning
Stars
241
Readiness
ready (97/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora meshcore openhop-core repeater sx1262

meshcore-dev/meshcore-ha

Home Assistant integration for monitoring and controlling MeshCore radio networks

Category
fine tuning
Stars
236
Readiness
ready (92/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +5 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ha homeassistant homeassistant-integration lora lorawan mesh

signetlabdei/lorawan

An ns-3 module for simulation of LoRaWAN networks

Category
fine tuning
Stars
222
Readiness
ready (76/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora lora-alliance lorawan network-analysis network-simulator simulation

KMX415/meshpoint

Open-source LoRa base station for Meshtastic/MeshCore. Raspberry Pi + SX1302/SX1303 concentrator: passive multi-channel capture, local dashboard, native TX

Category
fine tuning
Stars
218
Readiness
ready (82/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +4 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

helium helium-hotspot helium-mining hotspot iot-gateway lora

meshtastic/Meshtasticator

Discrete-event and interactive simulator for Meshtastic.

Category
fine tuning
Stars
205
Readiness
ready (76/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

discrete-event-simulation lora meshtastic network-simulation

jeremiah-k/meshtastic-matrix-relay

A relay between a Matrix room and a Meshtastic radio. This relay extends your Matrix.org-based communication with a LoRa-based Meshtastic radio mesh. This is not an official product of Matrix.org or M

Category
fine tuning
Stars
197
Readiness
ready (83/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora matrix mesh-networks meshtastic radio

wuwangzhang1216/abliterix

Automated alignment adjustment for LLMs — direct steering, LoRA, and MoE expert-granular abliteration, optimized via multi-objective Optuna TPE.

Category
fine tuning
Stars
189
Readiness
ready (87/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +4 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

abliteration alignment decensoring gemma llm lora

perfectgf/lora-dataset-studio

Self-hosted, one-tab workbench for the whole LoRA lifecycle: build Character / Concept / Style datasets (generate from a reference, scrape, or triage a huge dump in the Image Bank), curate, caption, c

Category
fine tuning
Stars
173
Readiness
ready (86/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +49 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-toolkit comfyui dataset flask lora react

meshtastic/protobufs

Protobuf definitions for the Meshtastic project

Category
fine tuning
Stars
171
Readiness
ready (83/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

firmware iot lora mesh mesh-networking meshtastic

DaoyuanLi2816/pairjudge

Pairwise LLM judges (A/B/tie): budget-aware multi-turn packing, position-bias correction, pseudo-label distillation. Generalized from the 4th-place (gold) solution to Kaggle LMSYS Chatbot Arena.

Category
fine tuning
Stars
169
Readiness
ready (82/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chatbot-arena gold-medal kaggle-competition kaggle-solution llm llm-as-judge

chirpstack/chirpstack-gateway-os

OpenWrt based gateway images including ChirpStack components.

Category
fine tuning
Stars
158
Readiness
ready (73/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chirpstack gateway iot lora lorawan openwrt

nopnop2002/esp-idf-sx126x

SX1262/SX1268/LLCC68 Low Power Long Range Transceiver driver for esp-idf

Category
fine tuning
Stars
151
Readiness
ready (82/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ebyte ebyte-e22 ebyte-e220 esp-idf esp32 llcc68

phi-monster/Galahad

Instruction blindness in vision-language-action policies: diagnosis and a low-rank data cure. Paper, model, deconfounded datasets, generator, and the measurement battery.

Category
fine tuning
Stars
151
Readiness
ready (86/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +41 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

benchmark causal-confusion embodied-ai instruction-following interpretability language-grounding

MeshAddicts/meshinfo

Realtime web UI to run against a Meshtastic regional or private mesh network.

Category
fine tuning
Stars
143
Readiness
ready (81/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora m mesh mesh-networks meshtastic meshtastic-python

shootthesound/Fizgig

Krea 2 & Klein 9B LoRA - LoKR Studio — train, profile, repair, and extract Krea 2 & Flux 2 Klein 9B LoRAs & LoKRs

Category
fine tuning
Stars
136
Readiness
ready (91/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +26 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai deep-learning flux flux2 generative-ai klein

dabeani/meshcoreterm

MeshCoreTerm (MC-Term); Retro Firmware based on the MeshCore SourceCode for LilyGo TDeck, Seeedstudio Indicator, Heltec V4, EleCrow 3.5, EleCrow 7

Category
fine tuning
Stars
128
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: strong signals despite limited visibility

Strongest signals: push recency, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

433mhz 868mhz 915mhz ble bluetooth elecrow

gantasmo/theDAW

Full-featured DAW, DJ, and VJ app interoperable w/Ableton, Reaper, Resolume & more. Stable Audio 3, Magenta RT2, Suno API, Chimera track fusion, Demucs stems, MIDI generate/notate, img > spectrogram >

Category
fine tuning
Stars
123
Readiness
ready (93/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-audio audio-inpainting audio-mixing daw dj ffmpeg

lorien/awesome-reticulum

A curated list of projects related to Reticulum Network

Category
fine tuning
Stars
123
Readiness
ready (94/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +11 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora lora-gateway lxmf lxmfy lxst mesh

GeneralDussDuss/poseidon

80+ feature pentesting firmware for M5Stack Cardputer-Adv. WiFi, BLE, sub-GHz (CC1101), 2.4GHz (nRF24), LoRa (SX1262), IR, BadUSB, DHCP attacks, WPAD, MouseJack, BLE spam, signal replay, 6 themes. Sup

Category
fine tuning
Stars
119
Readiness
needs review (69/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +6 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

arduino ble bluetooth cardputer cc1101 deauth

Tzohar/PassLLM

World's most accurate password guessing AI tool. A PyTorch implementation of PassLLM (USENIX 2025) that leverages PII and LoRA fine-tuning to outperform existing tools by over 45% on consumer hardware

Category
fine tuning
Stars
113
Readiness
ready (81/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai artificial-intelligence deep-learning generative-ai hacking hashcat

rockerBOO/lora-inspector

LoRA (Low-Rank Adaptation) inspector for Stable Diffusion

Category
fine tuning
Stars
100
Readiness
ready (87/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

analysis lora low-rank-approximation machine-learning pytorch

modelscope/mcore-bridge

MCore-Bridge: Providing Megatron-Core model definitions for state-of-the-art large models and making Megatron training as simple as Transformers — with support for 300+ large language models (Qwen3-Ne

Category
fine tuning
Stars
90
Readiness
ready (95/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

deepseek-r1 deepseek-v4 gemma4 glm-5 gpt-oss llama4

dada-bots/underfit

Stable Audio LoRA Trainer of salty goodness

Category
fine tuning
Stars
88
Readiness
ready (83/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +7 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

audio-ai lora lora-fine-tuning music-ai stability-ai stable-audio

luciobaiocchi/heard

Offline group-safety mesh for hikers: ESP32 + GPS + LoRa, with a firmware-in-the-loop simulator and 3D replay viewer

Category
fine tuning
Stars
85
Readiness
ready (78/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

digital-twin embedded esp32 freertos gps gps-tracking

lupyuen/lupyuen.github.io

Lup Yuen's Articles and Resume

Category
fine tuning
Stars
82
Readiness
ready (77/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

cortex embedded grafana iot lora lorawan

Techposts/TankSync

Solar-powered long-range water-tank monitor — LoRa to an ESP32 hub with OLED + LED indicators. MQTT-native + Home Assistant integration via HACS. Open firmware (AGPL-3.0), open hardware (CC BY-SA 4.0)

Category
fine tuning
Stars
80
Readiness
ready (91/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agpl esp-idf esp32 hacs home-assistant iot

NVlabs/LoRWeB

We propose a novel modular framework that learns to dynamically mix low-rank adapters (LoRAs) to improve visual analogy learning, enabling flexible and generalizable image edits based on example trans

Category
fine tuning
Stars
75
Readiness
ready (80/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

analogies image-analogies image-editing lora visual-analogies

vavo/lora-pilot

LoRA Pilot is an ultimate docker image for all Stable Diffusion LoRA trainers. Includes kohya_ss, diffusion pipes and TensorBoard for trainings and ComfyUI and InvokeAI for validation. Features shared

Category
fine tuning
Stars
74
Readiness
ready (85/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals; strong signals despite limited visibility

Strongest signals: push recency, issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-toolkit captioning-images comfyui datasets-preparation diffusion-pipe docker-image

icssw-org/MeshCom-Firmware

MeshCom Client 4.0

Category
fine tuning
Stars
74
Readiness
ready (97/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, issue load, fork interest. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

amateur-radio emergency-services lora mesh-networks resilience telemetry

labmlai/annotated_deep_learning_paper_implementations

🧑‍🏫 60+ Implementations/tutorials of deep learning papers with side-by-side notes 📝; including transformers (original, xl, switch, feedback, vit, ...), optimizers (adam, adabelief, sophia, ...), gans(

Category
fine tuning
Stars
67,291
Readiness
needs review (45/100 heuristic points; not a probability)
Data confidence
medium
Maintainer health
risky
Maintenance risk
16/100 · high confidence

Why: +42 stars in 7 days; 39 lifetime contributors

Why it may be a gem: open issue backlog is stable or shrinking

Strongest signals: contributor breadth, issue load, documentation. Risks: no pull-request review responses in 30 days, no maintainer response activity in 30 days. Missing inputs: release recency.

attention deep-learning deep-learning-tutorial gan literate-programming lora

microsoft/LoRA

Code for loralib, an implementation of "LoRA: Low-Rank Adaptation of Large Language Models"

Category
fine tuning
Stars
13,718
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +19 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 598 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

adaptation deberta deep-learning gpt-2 gpt-3 language-model

LianjiaTech/BELLE

BELLE: Be Everyone's Large Language model Engine（开源中文对话大模型）

Category
fine tuning
Stars
8,279
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 660 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bloom chinese-nlp gpt-evaluation gpt-q instruct-finetune instruct-gpt

Akegarasu/lora-scripts

SD-Trainer. LoRA & Dreambooth training scripts & GUI use kohya-ss's trainer, for diffusion model.

Category
fine tuning
Stars
6,093
Readiness
high risk (44/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 333 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

dreambooth finetune lora stable-diffusion

nunchaku-ai/nunchaku

[ICLR2025 Spotlight] SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models

Category
fine tuning
Stars
3,932
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
25/100 · low confidence

Why: +11 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 153 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

comfyui diffusion-models flux genai iclr iclr2025

mymusise/ChatGLM-Tuning

基于ChatGLM-6B + LoRA的Fintune方案

Category
fine tuning
Stars
3,742
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: repository is archived, no push in 986 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chatglm chatgpt lora peft

agentheroes/agentheroes

Generate, animate and schedule your AI characters 🤖

Category
fine tuning
Stars
3,363
Readiness
high risk (38/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 410 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent agents ai chatgpt generate image

adapter-hub/adapters

A Unified Library for Parameter-Efficient and Modular Transfer Learning

Category
fine tuning
Stars
2,826
Readiness
needs review (56/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
17/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 103 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

adapters bert lora natural-language-processing nlp parameter-efficient-learning

gioblu/PJON

PJON (Padded Jittering Operative Network) is an experimental, arduino-compatible, multi-master, multi-media network protocol.

Category
fine tuning
Stars
2,804
Readiness
needs review (45/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 259 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

arduino attiny esp32 esp8266 home-automation iot

PhoebusSi/Alpaca-CoT

We unified the interfaces of instruction-tuning data (e.g., CoT data), multiple LLMs and parameter-efficient methods (e.g., lora, p-tuning) together for easy use. We welcome open-source enthusiasts to

Category
fine tuning
Stars
2,791
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 969 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

alpaca chatglm chatgpt cot instruction-tuning llama

liucongg/ChatGLM-Finetuning

基于ChatGLM-6B、ChatGLM2-6B、ChatGLM3-6B模型，进行下游具体任务微调，涉及Freeze、Lora、P-tuning、全参微调等

Category
fine tuning
Stars
2,772
Readiness
high risk (42/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 969 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chatglm chatglm2 chatglm3 chatgpt freeze lora

JIA-Lab-research/LongLoRA

Code and documents of LongLoRA and LongAlpaca (ICLR 2024 Oral)

Category
fine tuning
Stars
2,689
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 723 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

fine-tuning-llm large-language-models llm long-context lora

wenge-research/YAYI

雅意大模型：为客户打造安全可靠的专属大模型，基于大规模中英文多领域指令数据训练的 LlaMA 2 & BLOOM 系列模型，由中科闻歌算法团队研发。(Repo for YaYi Chinese LLMs based on LlaMA2 & BLOOM)

Category
fine tuning
Stars
2,532
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: repository is archived, no push in 933 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bloom chat chinese llama llama2 llm

markqvist/NomadNet

Communicate Freely

Category
fine tuning
Stars
2,370
Readiness
high risk (44/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +34 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora lxmf mesh-networks off-grid packet-radio privacy

chenking2020/FindTheChatGPTer

ChatGPT爆火，开启了通往AGI的关键一步，本项目旨在汇总那些ChatGPT的开源平替们，包括文本大模型、多模态大模型等，为大家提供一些便利

Category
fine tuning
Stars
2,005
Readiness
high risk (42/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 1089 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agi alpaca autogpt baichuan belle ceval

siliconflow/onediff

OneDiff: An out-of-the-box acceleration library for diffusion models.

Category
fine tuning
Stars
1,964
Readiness
needs review (51/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 246 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

aigc-serving comfyui comfyui-workflow cuda diffusers diffusion-models

ssbuild/chatglm_finetuning

chatglm 6b finetuning and alpaca finetuning

Category
fine tuning
Stars
1,526
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 516 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

adalora chatglm deep-learning freeze ia3 lora

brocaar/chirpstack-network-server

ChirpStack Network Server is an open-source LoRaWAN network-server.

Category
fine tuning
Stars
1,514
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: repository is archived, no push in 462 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

internet-of-things iot lora lora-server loraserver lorawan

zjunlp/KnowLM

An Open-sourced Knowledgable Large Language Model Framework.

Category
fine tuning
Stars
1,387
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 573 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bilingual chinese deep-learning deepspeed english gpt-3

whitecatboard/Lua-RTOS-ESP32

Lua RTOS for ESP32

Category
fine tuning
Stars
1,326
Readiness
needs review (49/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
17/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: limited evidence; inspect maintenance signals before adopting

Strongest signals: issue load, fork interest, documentation. Risks: no push in 100 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

esp32 espressif freertos iot iot-platform lora

yeyupiaoling/Whisper-Finetune

Fine-tune the Whisper speech recognition model to support training without timestamp data, training with timestamp data, and training without speech data. Accelerate inference and support Web deployme

Category
fine tuning
Stars
1,221
Readiness
needs review (59/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
15/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 91 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

android asr chinese ctranslate2 huggingface lora

SkalskiP/vlms-zero-to-hero

This series will take you on a journey from the fundamentals of NLP and Computer Vision to the cutting edge of Vision-Language Models.

Category
fine tuning
Stars
1,178
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 561 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bert-model clip computer-vision embeddings gpt gpt-2

punica-ai/punica

Serving multiple LoRA finetuned LLM as one

Category
fine tuning
Stars
1,170
Readiness
needs review (47/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, license. Risks: no push in 821 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

large-language-models llm lora

liamcottle/reticulum-meshchat

A simple mesh network communications app powered by the Reticulum Network Stack.

Category
fine tuning
Stars
1,151
Readiness
needs review (70/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +5 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

codec2 ed25519 electron encryption lora lxmf

unit-mesh/unit-minions

《AI 研发提效：自己动手训练 LoRA》，包含 Llama （Alpaca LoRA）模型、ChatGLM （ChatGLM Tuning）相关 Lora 的训练。训练内容：用户故事生成、测试代码生成、代码辅助生成、文本转 SQL、文本生成代码……

Category
fine tuning
Stars
1,099
Readiness
high risk (36/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load. Risks: no push in 947 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

llm lora

gotthardp/lorawan-server

Compact server for private LoRaWAN networks

Category
fine tuning
Stars
1,004
Readiness
needs review (57/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 986 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

erlang lora lorawan lorawan-server

tapparelj/gr-lora_sdr

This is the fully-functional GNU Radio software-defined radio (SDR) implementation of a LoRa transceiver with all the necessary receiver components to operate correctly even at very low SNRs. This wor

Category
fine tuning
Stars
990
Readiness
high risk (40/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +5 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load. Risks: no push in 214 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora sdr

NVlabs/DoRA

[ICML2024 (Oral)] Official PyTorch implementation of DoRA: Weight-Decomposed Low-Rank Adaptation

Category
fine tuning
Stars
987
Readiness
high risk (44/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
23/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 136 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

commonsense-reasoning deep-learning deep-neural-networks instruction-tuning large-language-models large-vision-language-models

ElectronicCats/CatSniffer

CatSniffer is an original multiprotocol and multiband board for sniffing, communicating, and attacking IoT (Internet of Things) devices using the latest radio IoT protocols. It is a highly portable US

Category
fine tuning
Stars
907
Readiness
needs review (46/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
29/100 · low confidence

Why: +6 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 176 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ble hardware lora matter rp2040 samd21

WangRongsheng/ChatGenTitle

🌟 ChatGenTitle：使用百万arXiv论文信息在LLaMA模型上进行微调的论文题目生成模型

Category
fine tuning
Stars
835
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: repository is archived, no push in 1091 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

arxiv large-language-models llama llm llms lora

kale5195/chilloutai

AI 图片生成

Category
fine tuning
Stars
807
Readiness
high risk (41/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: limited evidence; inspect maintenance signals before adopting

Strongest signals: issue load, documentation. Risks: no push in 1185 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chilloutmix civitai lora runpod

X-D-Lab/MindChat

🐋MindChat（漫谈）：漫谈人生路, 笑对风霜途

Category
fine tuning
Stars
717
Readiness
needs review (45/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 693 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

baichuan-13b chatglm2-6b chatgpt domain-llm internlm large-language-models

buildwithparallel/haven-manet-ip-mesh-radio

Haven – MANET IP Mesh Radio toolkit

Category
fine tuning
Stars
674
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: +14 stars in 7 days

Why it may be a gem: limited evidence; inspect maintenance signals before adopting

Strongest signals: issue load, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

atak batman-adv decentralized halow lora mesh

markqvist/RNode_Firmware

RNode is an open, free and flexible digital radio interface with many uses

Category
fine tuning
Stars
654
Readiness
needs review (49/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
17/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest. Risks: no push in 105 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora mesh-networks radio

PlayVoice/lora-svc

singing voice change based on whisper, and lora for singing voice clone

Category
fine tuning
Stars
645
Readiness
needs review (55/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 1008 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora singing-voice-conversion speech-to-sing uni-svc vits vits-svc

GURPREETKAURJETHRA/END-TO-END-GENERATIVE-AI-PROJECTS

End to End Generative AI Industry Projects on LLM Models with Deployment_Awesome LLM Projects

Category
fine tuning
Stars
622
Readiness
needs review (62/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
23/100 · low confidence

Why: +5 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 560 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chainlit finetuning-llms gemini generative-ai gpt4o gradio-python-llm

timmbogner/Farm-Data-Relay-System

A system that uses ESP-NOW, LoRa, and other protocols to transport sensor data in remote areas without relying on WiFi.

Category
fine tuning
Stars
618
Readiness
needs review (60/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 190 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agriculture agritech agtech arduino esp-now esp32

jakkra/Mars-Rover

3D printed and driveable Curiosity/Perseverance inspired Rover

Category
fine tuning
Stars
589
Readiness
needs review (58/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 1378 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

3d-printing arduino curiosity-rover esp-idf esp32 lora

ukairia777/tensorflow-nlp-tutorial

tensorflow를 사용하여 텍스트 전처리부터, Topic Models, BERT, GPT, LLM과 같은 최신 모델의 다운스트림 태스크들을 정리한 Deep Learning NLP 저장소입니다.

Category
fine tuning
Stars
580
Readiness
needs review (50/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

bert bert-ner dpo huggingface keras-tutorial llama

shootthesound/comfyUI-Realtime-Lora

Train and block edit and save LoRAs directly inside ComfyUI for Z-image, Flux Klein, SDXL, Flux, WAN 2.2, SD 1.5

Category
fine tuning
Stars
543
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai comfyui flux klein lora lora-loader

brocaar/chirpstack-application-server

ChirpStack Application Server is an open-source LoRaWAN application-server.

Category
fine tuning
Stars
535
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: repository is archived, no push in 462 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

application-server chirpstack iot lora lorawan

meshtastic/network-management-client

A Meshtastic desktop client, allowing simple, offline deployment and administration of an ad-hoc mesh communication network. Built in Rust and TypeScript.

Category
fine tuning
Stars
474
Readiness
high risk (44/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 278 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

client desktop emergency-response lora mesh mesh-networks

zetavg/LLaMA-LoRA-Tuner

UI tool for fine-tuning and testing your own LoRA models base on LLaMA, GPT-J and more. One-click run on Google Colab. + A Gradio ChatGPT-like Chat UI to demonstrate your language models.

Category
fine tuning
Stars
472
Readiness
needs review (47/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 1166 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai alpaca alpaca-lora google-colab gpt gpt-j

TheThingsArchive/ttn

The Things Network Stack V2

Category
fine tuning
Stars
462
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: repository is archived, no push in 1712 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

golang internet-of-things iot lora lora-server lorawan

airaria/Visual-Chinese-LLaMA-Alpaca

多模态中文LLaMA&Alpaca大语言模型（VisualCLA）

Category
fine tuning
Stars
461
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 1107 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

alpaca chinese llama llm lora multimodal

yuanjie-ai/ChatLLM

轻松玩转LLM兼容openai&langchain，支持文心一言、讯飞星火、腾讯混元、智谱ChatGLM等

Category
fine tuning
Stars
449
Readiness
needs review (56/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 683 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

chatbase chatdb chatdoc chatgpt chatkb chatkg

Saquib764/omini-kontext

An inference and training framework for multiple image input in Flux Kontext dev

Category
fine tuning
Stars
444
Readiness
high risk (35/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load. Risks: no push in 340 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

artificial-intelligence kontext lora

xreef/LoRa_E32_Series_Library

Arduino LoRa EBYTE E32 device library complete and tested with Arduino, esp8266, esp32, STM32 and Raspberry Pi Pico (rp2040 boards). sx1278/sx1276

Category
fine tuning
Stars
426
Readiness
needs review (49/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 188 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

arduino arduino-library arduino-mkr arduino-nano-33-iot e32 ebyte

datawhalechina/llm-deploy

大模型/LLM推理和部署理论与实践

Category
fine tuning
Stars
416
Readiness
high risk (43/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: limited evidence; inspect maintenance signals before adopting

Strongest signals: issue load, documentation. Risks: no push in 389 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

knowledge-distillation llm llm-deploy lora pruning quantization

yangjianxin1/Firefly-LLaMA2-Chinese

Firefly中文LLaMA-2大模型，支持增量预训练Baichuan2、Llama2、Llama、Falcon、Qwen、Baichuan、InternLM、Bloom等大模型

Category
fine tuning
Stars
415
Readiness
high risk (40/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 1021 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

baichaun2 baichuan baichuan-13b bloom chatglm falcon

ExpressLRS/ExpressLRS-Hardware

Open-source PCB designs, schematics, and 3D-printable cases for ExpressLRS TX modules and receivers

Category
fine tuning
Stars
403
Readiness
needs review (49/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +3 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 300 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

diy drone esp32 esp8266 fpv lora

Hon-Wong/VoRA

[Fully open] [Encoder-free MLLM] Vision as LoRA

Category
fine tuning
Stars
389
Readiness
high risk (41/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 421 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

llm lora lvlm mllm vlm

lora-aprs/LoRa_APRS_iGate

This is a LoRa APRS iGate/Digi based on an ESP32

Category
fine tuning
Stars
389
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: repository is archived, no push in 327 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

aprs aprs-beacon aprs-igate aprs-is aprs-tracker arduino

km1994/llms_paper

该仓库主要记录 LLMs 算法工程师相关的顶会论文研读笔记（多模态、PEFT、小样本QA问答、RAG、LMMs可解释性、Agents、CoT）

Category
fine tuning
Stars
388
Readiness
high risk (39/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 862 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

agent llms lora peft qa rag

TUDB-Labs/mLoRA

An Efficient "Factory" to Build Multiple LoRA Adapters

Category
fine tuning
Stars
383
Readiness
needs review (58/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 540 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

baichuan chatglm dpo finetune gpu llama

batear-io/batear

● batear — Ultra-low-cost, off-grid acoustic drone detector. Edge computing. Protect your airspace with a microphone.

Category
fine tuning
Stars
381
Readiness
needs review (56/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
16/100 · low confidence

Why: +9 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 96 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

acoustic anti-war drone drone-detection esp32 heltec

lora-aprs/LoRa_APRS_Tracker

A LoRa based APRS tracker for ESP32 boards.

Category
fine tuning
Stars
367
Readiness
high risk (0/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
100/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: repository is archived, no push in 327 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

aprs aprs-beacon aprs-igate aprs-is aprs-tracker arduino

LilyGO/TTGO-T-Beam

No description

Category
fine tuning
Stars
358
Readiness
high risk (41/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: limited evidence; inspect maintenance signals before adopting

Strongest signals: issue load, fork interest. Risks: no push in 1749 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

esp32-arduino lora

SMotlaq/LoRa

Ra-02 LoRa module (SX1278) library for STM32 (ARM processors) using HAL drivers ⚡

Category
fine tuning
Stars
326
Readiness
needs review (59/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 922 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

arm embedded iot lora lorawan microcontroller

p1atdev/LECO

Low-rank adaptation for Erasing COncepts from diffusion models.

Category
fine tuning
Stars
323
Readiness
needs review (48/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, license. Risks: no push in 930 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

diffusion lora stable-diffusion

liamcottle/meshtastic-map

A map of all Meshtastic nodes heard via MQTT.

Category
fine tuning
Stars
306
Readiness
needs review (61/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +3 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

lora map mesh meshtastic mqtt

sh123/codec2_talkie

Turn your Android phone into Amateur Radio Codec2/OPUS APRS enabled DV handheld transceiver (Bluetooth/BLE/USB/TCPIP KISS/Sound modem client for DV digital voice communication)

Category
fine tuning
Stars
288
Readiness
needs review (49/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
27/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 164 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

amateur-radio amateurradio aprs bluetooth codec2 digital

EricLBuehler/xlora

X-LoRA: Mixture of LoRA Experts

Category
fine tuning
Stars
284
Readiness
needs review (50/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 734 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

llm lora python pytorch

meshtastic/antenna-reports

Community-contributed antenna testing reports and evaluations for Meshtastic devices.

Category
fine tuning
Stars
282
Readiness
high risk (42/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 478 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

antenna antennas lora meshtastic

sh123/esp32_loraprs

LoRa ESP32 KISS Bluetooth modem (for APRSDroid or aprs.fi iOS) + APRS-IS RX/TX iGate over WiFi + Digipeater + DV (with Codec2 Walkie-Talkie)

Category
fine tuning
Stars
272
Readiness
needs review (49/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
21/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation. Risks: no push in 128 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

amateur-radio aprs aprs-rx aprs-tracker aprsdroid aprsis

xyfJASON/ctrlora

[ICLR 2025] Codebase for "CtrLoRA: An Extensible and Efficient Framework for Controllable Image Generation"

Category
fine tuning
Stars
269
Readiness
needs review (48/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
26/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 154 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

controllable-generation controlnet image-to-image lora

KrisKasprzak/EBYTE

Libraries to program and use UART-based EBYTE wireless data transceivers

Category
fine tuning
Stars
262
Readiness
needs review (50/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, fork interest, documentation. Risks: no push in 727 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

arduino e32 e44 e44-ttl e50 e50-ttl

mlpc-ucsd/BLIVA

(AAAI 2024) BLIVA: A Simple Multimodal LLM for Better Handling of Text-rich Visual Questions

Category
fine tuning
Stars
261
Readiness
needs review (53/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 845 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

blip2 bliva chatbot instruction-tuning llama llm

tsiendragon/qwen-image-finetune

Repo for Qwen Image Finetune

Category
fine tuning
Stars
251
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
0/100 · low confidence

Why: +1 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

diffusion-models flux-kontext fsdp image-edit-model image-to-image lora

siara-cc/Unishox2

Compression for Unicode short strings (works on arduino)

Category
fine tuning
Stars
245
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 209 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

arduino bandwidth-saver chat-message-compression cloud-cost-intelligence compression cost-optimization

DSXiangLi/DecryptPrompt

总结Prompt&LLM论文，开源数据&模型，AIGC应用

Category
fine tuning
Stars
3,430
Readiness
high risk (42/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
watch
Maintenance risk
16/100 · low confidence

Why: +8 stars in 30 days

Why it may be a gem: limited evidence; inspect maintenance signals before adopting

Strongest signals: issue load, documentation. Risks: no push in 94 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

aigc chain-of-thought chatgpt demonstration few-shot-learning in-context-learning

ContextualAI/gritlm

Generative Representational Instruction Tuning

Category
fine tuning
Stars
697
Readiness
needs review (52/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +2 stars in 30 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 408 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

embedding embedding-models embeddings grit information-retrieval instruction-tuning

zyds/transformers-code

手把手带你实战 Huggingface Transformers 课程视频同步更新在B站与YouTube

Category
fine tuning
Stars
4,047
Readiness
high risk (39/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
30/100 · low confidence

Why: +3 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load. Risks: no push in 753 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

huggingface peft transformers

HuaizhengZhang/AI-Infra-from-Zero-to-Hero

🚀 Awesome System for Machine Learning ⚡️ AI System Papers and Industry Practice. ⚡️ System for Machine Learning, LLM (Large Language Model), GenAI (Generative AI). 🍻 OSDI, NSDI, SIGCOMM, SoCC, MLSys,

Category
fine tuning
Stars
4,267
Readiness
needs review (54/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
risky
Maintenance risk
16/100 · low confidence

Why: +22 stars in 7 days

Why it may be a gem: healthy maintenance and project fundamentals

Strongest signals: issue load, documentation, license. Risks: no push in 378 days. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-infra genai large-language-models llmsys mlsys model-serving

zhangjiayang6835-cyber/ai-research

🍯 自进化 AI 能力研究平台：蜜罐系统 → 沙箱评测引擎 → 训练数据集与 LoRA 微调流水线 / Self-Evolving AI Security Research Platform

Category
fine tuning
Stars
13
Readiness
ready (81/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, fork interest, documentation. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

ai-security anti-cheat code-analysis docker fine-tuning honeypot

TensorHarmony/lorakit

Fast, config-driven SDXL LoRA & DreamBooth fine-tuning — train from a single YAML file. QLoRA + torch.compile.

Category
fine tuning
Stars
19
Readiness
ready (80/100 heuristic points; not a probability)
Data confidence
low
Maintainer health
healthy
Maintenance risk
0/100 · low confidence

Why: High-signal fine tuning project

Why it may be a gem: strong signals despite limited visibility; healthy maintenance and project fundamentals

Strongest signals: push recency, documentation, license. Risks: None identified. Missing inputs: commit activity, contributor breadth, release recency, response activity, maintenance distribution.

diffusers dreambooth fine-tuning generative-ai lora pytorch