litgpt

20+ high-performance LLMs with recipes to pretrain, finetune and deploy at scale.

View on GitHub↗Upstream Lightning-AI/litgpt↗

Builder

Lightning-AI

Lightning-AI • individual

Stars

13,274

Using upstream star count

Forks

1,412

Using upstream fork count

Open Issues

Activity Score

0/100

5 commits in 30d

Created

May 4, 2023

Project creation date

README Summary

LitGPT is a comprehensive toolkit providing 20+ high-performance large language models with production-ready recipes for pretraining, finetuning, and deployment at scale. The repository offers optimized implementations and workflows for working with state-of-the-art LLMs in research and production environments. It focuses on performance, scalability, and ease of use for machine learning practitioners.

AI Dev Skills

Unmapped

Tags

Large Language Model TrainingTransformer Architecture ImplementationLoRA Fine-tuningQLoRA Fine-tuningModel QuantizationDistributed TrainingParameter-Efficient Fine-tuningModel OptimizationRLHF (Reinforcement Learning from Human Feedback)Model CompressionGPU Memory OptimizationPyTorch Lightning FrameworkModel Serving and DeploymentEfficient Model DeploymentOpen Source LLMsSelf-hostedEfficient AI TrainingLarge Language Model PretrainingOn-premiseModel Performance OptimizationCloudProduction LLM ServingCustom LLM Fine-tuningEdgeOn-device AILarge Language ModelsModel Adaptation for Specific DomainsResearch and Experimentation with LLMsTextPython

Recent Activity

Updated 22 days ago

7 Days

30 Days

90 Days

Quality

production

Quality: high
Maturity: production

PM Skills

Product Discovery

Languages

Python100.0%

Timeline

Project created: May 4, 2023
Forked: Mar 23, 2026
Your last push: 22 days ago
Upstream last push: 7 days ago
Tracked since: Mar 22, 2026

Similar Repos

pgvector cosine similarity · $0

Loading…