function: llm-training
818 products, primary matches first, then adoption-weighted; health v2 shown.
| Product | Health v2 | Stars | Maturity |
|---|---|---|---|
| deepset-ai/FARM FARM is a Python framework for fine-tuning and evaluating transformer-based language models for NLP tasks, with a focus on question answeri… | 10 | 1752 | maintenance |
| airaria/TextBrewer TextBrewer is a PyTorch-based knowledge distillation toolkit for NLP models. It provides an easy-to-use framework implementing state-of-the… | 23 | 1708 | maintenance |
| imcaspar/gpt2-ml GPT2-ML is a TensorFlow-based GPT-2 training and inference project supporting multiple languages, with a focus on Chinese. It provides 1.5B… | 23 | 1701 | maintenance |
| microsoft/EdgeML A library from Microsoft Research India implementing resource-efficient machine learning algorithms (Bonsai, ProtoNN, FastGRNN, EMI-RNN, DR… | 23 | 1681 | maintenance |
| tensorflow/mesh Mesh TensorFlow is a Python library and embedded language for specifying distributed tensor computations, letting users define how model di… | 10 | 1630 | maintenance |
| timoschick/pet Official implementation of Pattern-Exploiting Training (PET), a semi-supervised method that reformulates text classification and natural la… | 32 | 1622 | maintenance |
| lucidrains/PaLM-rlhf-pytorch A PyTorch library implementing Reinforcement Learning from Human Feedback (RLHF) on top of the PaLM transformer architecture, aiming to rep… | 78 | 7868 | experimental |
| Kotlin/kotlindl KotlinDL is a high-level deep learning framework written in Kotlin and inspired by Keras, built on top of the TensorFlow Java API and ONNX … | 23 | 1576 | maintenance |
| tanluren/yolov3-channel-and-layer-pruning A Python toolkit built on ultralytics/yolov3 that implements channel pruning, layer pruning, and knowledge distillation for YOLOv3/v4 (incl… | 32 | 1515 | maintenance |
| open-mmlab/Multimodal-GPT Multimodal-GPT is an open-source project for training a multimodal chatbot that combines vision and language instructions, built on OpenFla… | 30 | 1512 | maintenance |
| maderix/ANE A research project demonstrating backpropagation and neural network training directly on Apple's Neural Engine using reverse-engineered pri… | 47 | 7253 | experimental |
| jina-ai/finetuner Jina AI's Finetuner is a Python library for task-oriented fine-tuning of pretrained models like BERT and CLIP to produce better embeddings … | 10 | 1503 | maintenance |
| lxztju/pytorch_classification A complete PyTorch image classification codebase built on torchvision, covering training, prediction, TTA, model ensembling, knowledge dist… | 32 | 1466 | maintenance |
| CStanKonrad/long_llama LongLLaMA is a large language model handling long contexts of 256k tokens or more, built on OpenLLaMA and fine-tuned with the Focused Trans… | 29 | 1465 | maintenance |
| BigScience BigScience is a research project training large transformer language models (BERT, GPT-style) at scale, built on a fork of Megatron-LM inte… | 32 | 1448 | maintenance |
| mit-han-lab/proxylessnas ProxylessNAS is a neural architecture search (NAS) framework that directly searches CNN architectures on the target task and target hardwar… | 32 | 1447 | maintenance |
| OpenLMLab/MOSS-RLHF MOSS-RLHF is the open-source companion code for the paper 'Secrets of RLHF in Large Language Models Part I: PPO', providing implementations… | 29 | 1429 | maintenance |
| instructlab/instructlab InstructLab (ilab) is an open-source CLI and Python package for enhancing large language models through community-contributed taxonomy-base… | 10 | 1419 | maintenance |
| thunlp/ERNIE ERNIE is a PyTorch toolkit and dataset for augmenting pre-trained language models like BERT with knowledge graph entity representations, fr… | 32 | 1417 | maintenance |
| ConnorJL/GPT2 A community Python/TensorFlow implementation of GPT-2 model training and text generation that supports both GPUs and TPUs. It includes scri… | 32 | 1412 | maintenance |
| JonasGeiping/cramming A research framework for pretraining BERT-type language models from scratch on a single consumer GPU in one day, replicating the paper 'Cra… | 22 | 1366 | maintenance |
| hiyouga/FastEdit FastEdit is a Python library and CLI tool for editing large language models, injecting fresh or customized factual knowledge into pretraine… | 19 | 1366 | maintenance |
| facebookresearch/moco-v3 A PyTorch implementation of MoCo v3, a self-supervised contrastive learning method for ResNet and Vision Transformer (ViT) models. It inclu… | 10 | 1323 | maintenance |
| FreedomIntelligence/HuatuoGPT HuatuoGPT is an open Chinese medical large language model (7B and 13B variants) fine-tuned on a hybrid of distilled and real-world medical … | 30 | 1322 | maintenance |
| 920232796/bert_seq2seq A lightweight PyTorch framework for fine-tuning pretrained language models (BERT, RoBERTa, Nezha, GPT2, T5, BART) on Chinese NLP tasks usin… | 32 | 1308 | maintenance |
| uclaml/SPIN Official implementation of Self-Play Fine-Tuning (SPIN), a method that improves language models by having them iteratively generate and dis… | 26 | 1254 | maintenance |
| kamalkraj/BERT-NER A PyTorch library for training and running named entity recognition (NER) models based on Google's BERT, evaluated on the CoNLL-2003 datase… | 32 | 1249 | maintenance |
| KMnP/vpt Official PyTorch implementation of Visual Prompt Tuning (VPT), an ECCV 2022 method for parameter-efficient fine-tuning of vision transforme… | 32 | 1244 | maintenance |
| AGI-Edgerunners/LLM-Adapters LLM-Adapters is a Python framework extending HuggingFace's PEFT library that integrates various adapter types (LoRA, series/parallel adapte… | 30 | 1233 | maintenance |
| google-research/fixmatch Official research code for FixMatch, a semi-supervised learning method that combines consistency regularization with confidence-based pseud… | 10 | 1223 | maintenance |
| awslabs/sockeye Sockeye is an open-source sequence-to-sequence framework for Neural Machine Translation built on PyTorch, with distributed training and opt… | 23 | 1215 | maintenance |
| UMass-Embodied-AGI/3D-LLM 3D-LLM is the research code for a large language model that takes 3D representations (objects and scenes) as input, built on BLIP-2/LAVIS. … | 28 | 1212 | maintenance |
| mit-han-lab/tinyml MIT Han Lab's TinyML research repository containing projects like TinyTL and NetAug for memory-efficient deep learning on microcontrollers … | 32 | 1211 | maintenance |
| henrywoo/chatllama An open-source Python library implementing RLHF-based fine-tuning of Meta's LLaMA models to build ChatGPT-style chat assistants, runnable o… | 22 | 1201 | maintenance |
| abhishekkrthakur/tez Tez is a lightweight, simple trainer library for PyTorch that simplifies training loops while keeping users close to native PyTorch. It sup… | 23 | 1155 | maintenance |
| IBM/Dromedary Dromedary is an open-source self-aligned language model trained with minimal human supervision using the principle-driven SELF-ALIGN pipeli… | 49 | 1137 | maintenance |
| microsoft/ToRA ToRA is a series of tool-integrated reasoning LLM agents from Microsoft that solve mathematical reasoning problems by interleaving natural … | 27 | 1124 | maintenance |
| microsoft/MASS MASS is Microsoft's PyTorch implementation of Masked Sequence to Sequence Pre-training for language generation tasks. It provides pre-train… | 10 | 1115 | maintenance |
| keroro824/HashingDeepLearning SLIDE is a C++ research codebase implementing locality-sensitive hashing based training of deep neural networks, from the paper 'In Defense… | 32 | 1103 | maintenance |
| EleutherAI/math-lm Repository for Llemma, an open language model for mathematics, hosting training code, data preprocessing scripts, and evaluation tooling. I… | 31 | 1097 | maintenance |
| RUCAIBox/TextBox TextBox 2.0 is a Python/PyTorch library providing a unified pipeline for applying pre-trained language models to text generation tasks. It … | 23 | 1097 | maintenance |
| Tencent/TencentPretrain TencentPretrain is a PyTorch-based toolkit for pre-training and fine-tuning transformer models across modalities such as text, vision, and … | 32 | 1091 | maintenance |
| replit/ReplitLM Official repository with inference code, configs, and guides for the ReplitLM family of code-oriented language models, such as replit-code-… | 10 | 1082 | maintenance |
| MatthieuCourbariaux/BinaryNet BinaryNet is a research codebase for training deep neural networks whose weights and activations are constrained to +1 or -1, reproducing t… | 32 | 1068 | maintenance |
| jondurbin/airoboros Airoboros is a Python library implementing a heavily modified version of the Self-Instruct paper to generate high-quality synthetic instruc… | 20 | 1051 | maintenance |
| yangxudong/deeplearning A collection of deep learning model training, evaluation, and prediction code implemented with TensorFlow's high-level Estimator API, with … | 32 | 1049 | maintenance |
| Xwin-LM/Xwin-LM Xwin-LM is an open-source project for LLM alignment technologies including supervised fine-tuning, reward models, reject sampling, and RLHF… | 28 | 1035 | maintenance |
| dandelionsllm/pandallm Panda is an open-source project for overseas Chinese large language models, providing PandaLLM model weights (continued pretraining of LLaM… | 30 | 1031 | maintenance |
| cxxnet ps-lite is a lightweight, efficient C++ implementation of the parameter server framework for distributed machine learning. It exposes simpl… | 10 | 1024 | maintenance |
| tensorflow/neural-structured-learning Neural Structured Learning (NSL) is a TensorFlow framework for training neural networks with structured signals, either explicit graphs or … | 65 | 1010 | maintenance |
| llSourcell/Doctor-Dignity Doctor Dignity is a fine-tuned Llama2 7B model that can pass the US Medical Licensing Exam, built with PyTorch, Transformers, TRL, and ONNX… | 28 | 3822 | experimental |
| MoonshotAI/Attention-Residuals Official implementation of Attention Residuals (AttnRes), a drop-in replacement for standard residual connections in Transformers that lets… | 47 | 3487 | experimental |
| xjdr-alt/entropix Entropix is a research project implementing entropy-based sampling and parallel chain-of-thought decoding for large language models, aiming… | 22 | 3432 | experimental |
| Everlyn-Labs/Everlyn-1 Everlyn-1 is an open autoregressive foundational video AI model from Everlyn Labs, accompanied by research on video compression/tokenizatio… | 22 | 2892 | experimental |
| openai/weak-to-strong OpenAI's research codebase implementing weak-to-strong generalization experiments from their alignment paper, where a strong pretrained mod… | 10 | 2552 | experimental |
| hyperspaceai/agi Hyperspace AGI is an experimental peer-to-peer network where autonomous AI agents collaboratively train language models and share research … | 82 | 2031 | experimental |
| Continual-Intelligence/SEAL SEAL (Self-Adapting LLMs) is a research framework from MIT CSAIL that trains language models via reinforcement learning to generate their o… | 34 | 1850 | experimental |
| EvolvingLMMs-Lab/open-r1-multimodal A fork of Hugging Face's open-r1 that extends the R1 GRPO reinforcement learning training paradigm to multimodal (vision-language) models l… | 23 | 1603 | experimental |
| lyuchenyang/Macaw-LLM Macaw-LLM is a multi-modal language modeling framework that integrates image, video, audio, and text data, built on CLIP, Whisper, and LLaM… | 29 | 1591 | experimental |
| AnswerDotAI/fsdp_qlora A training script/library from Answer.AI that combines QLoRA (quantized LoRA) with PyTorch FSDP to fine-tune large language models like Lla… | 26 | 1550 | experimental |
| KhoomeiK/LlamaGym LlamaGym is a Python library that simplifies fine-tuning LLM-based agents with online reinforcement learning in Gym-style environments. It … | 25 | 1254 | experimental |
| SakanaAI/self-adaptive-llms Transformer² is a research framework from SakanaAI that adapts large language models to unseen tasks in real-time by selectively adjusting … | 23 | 1225 | experimental |
| Jamie-Stirling/RetNet A minimal, pure PyTorch implementation of the Retentive Network (RetNet) architecture proposed as a successor to Transformers for large lan… | 28 | 1209 | experimental |
| danielgross/LlamaAcademy LlamaAcademy is a Python pipeline that crawls API documentation, generates synthetic training data with GPT-3.5/GPT-4, and fine-tunes a Vic… | 30 | 1198 | experimental |
| AutoArk/TinyEngram TinyEngram is an open research project exploring DeepSeek-AI's Engram architecture and memory injection as an alternative to LoRA for param… | 53 | 1191 | experimental |
| facebookresearch/shumai Shumai is a fast, differentiable tensor library for TypeScript and JavaScript built on Bun and Flashlight (ArrayFire backend). It provides … | 32 | 1172 | experimental |
| kijai/ComfyUI-FluxTrainer A ComfyUI custom node plugin that wraps kohya's sd-scripts to enable LoRA, LyCORIS, and full fine-tune training of FLUX models directly ins… | 29 | 1160 | experimental |
| Ligo-Biosciences/AlphaFold3 An open-source Python implementation of AlphaFold3, DeepMind's biomolecular structure prediction model, including the full model architectu… | 27 | 1091 | experimental |
| volcengine/veScale veScale is a PyTorch distributed training library from ByteDance for hyperscale training of large language models and reinforcement learnin… | 57 | 1036 | experimental |
| HumanMLLM/R1-Omni R1-Omni is a research project applying Reinforcement Learning with Verifiable Reward (RLVR) to an omni-multimodal large language model for … | 26 | 1022 | experimental |
| LAION-AI/Open-Assistant OpenAssistant is an open-source chat-based assistant that understands tasks, interacts with third-party systems, and retrieves information … | 22 | 37408 | abandoned |
| horovod/horovod Horovod is a distributed deep learning training framework for TensorFlow, Keras, PyTorch, and Apache MXNet, originally developed at Uber. I… | 10 | 14688 | abandoned |
| Jiayi-Pan/TinyZero TinyZero is a minimal reproduction of DeepSeek R1-Zero, showing that a 3B base language model can develop self-verification and search abil… | 52 | 13223 | abandoned |
| intel/ipex-llm IPEX-LLM is a PyTorch LLM acceleration library for Intel hardware (iGPU, NPU, Arc/Flex/Max GPUs, and CPU), offering low-bit quantization (F… | 10 | 8859 | abandoned |
| facebookarchive/caffe2 Caffe2 is a lightweight, modular, and scalable deep learning framework built on the original Caffe, with both Python and C++ APIs. It has b… | 10 | 8370 | abandoned |
| nebuly-ai/optimate OptiMate is a collection of Python libraries from Nebuly AI for optimizing AI model performance, including Speedster for inference accelera… | 23 | 8329 | abandoned |
| EleutherAI/gpt-neo GPT-Neo is EleutherAI's implementation of model- and data-parallel GPT-3-style transformer language models built on mesh-tensorflow, with r… | 10 | 8270 | abandoned |
| Lightning-AI/lit-llama Lit-LLaMA is an independent, Apache 2.0-licensed implementation of the LLaMA language model built on nanoGPT, covering pre-training, fine-t… | 43 | 6085 | abandoned |
| huggingface/autotrain-advanced AutoTrain Advanced is Hugging Face's no-code tool for training and deploying state-of-the-art machine learning models, covering LLM finetun… | 74 | 4608 | abandoned |
| facebookresearch/fairseq-lua Fairseq-lua is Facebook AI Research's sequence-to-sequence learning toolkit for the Torch framework, focused on neural machine translation … | 10 | 3725 | abandoned |
| hiyouga/ChatGLM-Efficient-Tuning A Python framework for parameter-efficient fine-tuning (LoRA, QLoRA, P-Tuning, RLHF) of the ChatGLM-6B and ChatGLM2-6B language models usin… | 10 | 3713 | abandoned |
| latitudegames/AIDungeon AI Dungeon 2 is an open-source, infinitely generated text adventure game powered by a fine-tuned GPT-2 language model. It can be played loc… | 10 | 3224 | abandoned |
| alpa-projects/alpa Alpa is a Python system for training and serving large-scale neural networks by automatically parallelizing single-device code across distr… | 10 | 3178 | abandoned |
| mistralai/mistral-finetune A lightweight Python codebase from Mistral AI for memory-efficient LoRA fine-tuning of Mistral's language models, optimized for single-node… | 10 | 3096 | abandoned |
| sony/nnabla Sony's Neural Network Libraries (nnabla) is a deep learning framework with a Python API over a C++11 core, designed for research, developme… | 10 | 2774 | abandoned |
| microsoft/DMTK DMTK is Microsoft's Distributed Machine Learning Toolkit, an umbrella project hosting a parameter server framework (Multiverso) plus distri… | 10 | 2738 | abandoned |
| intel/BigDL BigDL is Intel's distributed deep learning library that scales TensorFlow, Keras, and PyTorch workloads on Apache Spark, Flink, and Ray, wi… | 10 | 2698 | abandoned |
| neuralmagic/sparseml SparseML is a Python library for applying sparsification recipes (pruning, quantization, sparsity) to neural networks with a few lines of c… | 10 | 2145 | abandoned |
| openai/image-gpt OpenAI's official code and pre-trained models for iGPT (image GPT), a GPT-2-style transformer adapted to generate and classify images as pi… | 10 | 2096 | abandoned |
| lxe/simple-llm-finetuner A beginner-friendly Gradio web UI for fine-tuning LLMs (LLaMA, GPT-2) using LoRA via the Hugging Face PEFT library on consumer NVIDIA GPUs.… | 30 | 2054 | abandoned |
| nyu-mll/jiant jiant is a PyTorch-based NLP research toolkit for multitask and transfer learning, supporting 50+ natural language understanding tasks and … | 23 | 1673 | abandoned |
| huggingface/pytorch-openai-transformer-lm A PyTorch reimplementation of OpenAI's finetuned transformer language model (GPT-1) from the paper 'Improving Language Understanding by Gen… | 32 | 1522 | abandoned |
| openai/lm-human-preferences OpenAI's research code for the paper 'Fine-Tuning Language Models from Human Preferences', implementing reward model training from human la… | 10 | 1391 | abandoned |
| NousResearch/atropos Atropos is a Python framework from Nous Research for building reinforcement learning environments that collect and evaluate LLM trajectorie… | 10 | 1350 | abandoned |
| neuronika/neuronika Neuronika is a machine learning framework written in pure Rust providing tensors and dynamic neural networks with reverse-mode automatic di… | 32 | 1086 | abandoned |
| NVIDIA/sentiment-discovery A deprecated PyTorch codebase from NVIDIA for large-scale unsupervised language model pretraining and transfer to sentiment and emotion cla… | 23 | 1065 | abandoned |
| tensorflow/tfjs-node tfjs-node was the Node.js native binding for TensorFlow.js, providing accelerated training and inference of ML models in JavaScript server … | 10 | 1055 | abandoned |
| huggingface/transformers Hugging Face Transformers is a Python library that serves as the model-definition framework for state-of-the-art machine learning models ac… | 95 | 164475 | stable |
| deepseek-ai/DeepSeek-R1 DeepSeek-R1 is a family of open-weight large language models trained with large-scale reinforcement learning for reasoning, including DeepS… | 24 | 92038 | active |
| RVC-Boss/GPT-SoVITS GPT-SoVITS is a Python-based few-shot voice cloning and text-to-speech system with an integrated WebUI. It supports zero-shot TTS from a 5-… | 68 | 61255 | active |