Practical course about Large Language Models.

[SIGIR'24] The official implementation code of MOELoRA.

Official implementation of "DoRA: Weight-Decomposed Low-Rank Adaptation"

Code for NOLA, an implementation of "nola: Compressing LoRA using Linear Combination of Random Basis"

An Efficient LLM Fine-Tuning Factory Optimized for MoE PEFT

[ICML'24 Oral] APT: Adaptive Pruning and Tuning Pretrained Language Models for Efficient Training and Inference

Official code implemtation of paper AntGPT: Can Large Language Models Help Long-term Action Anticipation from Videos?

memory-efficient fine-tuning; support 24G GPU memory fine-tuning 7B

CRE-LLM: A Domain-Specific Chinese Relation Extraction Framework with Fine-tuned Large Language Model

High Quality Image Generation Model - Powered with NVIDIA A100

Mistral and Mixtral (MoE) from scratch

Fine-tune StarCoder2-3b for SQL tasks on limited resources with LORA. LORA reduces model size for faster training on smaller datasets. StarCoder2 is a family of code generation models (3B, 7B, and 15B), trained on 600+ programming languages from The Stack v2 and some natural language text such as Wikipedia, Arxiv, and GitHub issues.

PEFT is a wonderful tool that enables training a very large model in a low resource environment. Quantization and PEFT will enable widespread adoption of LLM.

Finetuning Large Language Models

In this repo I will share different topics on anything I want to know in nlp and llms

Test results of Kanarya and Trendyol models with and without fine-tuning techniques on the Turkish tweet hate speech detection dataset.

Fine-tuning Llama3 8b to generate JSON formats for arithmetic questions and process the output to perform calculations.

practical projects using LLM, VLM and Diffusion models

NTU Deep Learning for Computer Vision 2023 course