tomaarsen / attention_sinksLinks
Extend existing LLMs way beyond the original training length with constant memory usage, without retraining
☆733Updated last year
Alternatives and similar repositories for attention_sinks
Users that are interested in attention_sinks are comparing it to the libraries listed below
Sorting:
- Code for fine-tuning Platypus fam LLMs using LoRA☆631Updated last year
- [ICML'24 Spotlight] LLM Maybe LongLM: Self-Extend LLM Context Window Without Tuning☆662Updated last year
- ☆416Updated 2 years ago
- Landmark Attention: Random-Access Infinite Context Length for Transformers☆427Updated 2 years ago
- ☆583Updated last year
- This repository contains code and tooling for the Abacus.AI LLM Context Expansion project. Also included are evaluation scripts and bench…☆598Updated 2 years ago
- Official repository for LongChat and LongEval☆533Updated last year
- A bagel, with everything.☆325Updated last year
- YaRN: Efficient Context Window Extension of Large Language Models☆1,652Updated last year
- [ACL2023] We introduce LLM-Blender, an innovative ensembling framework to attain consistently superior performance by leveraging the dive…☆973Updated last year
- Tune any FALCON in 4-bit☆465Updated 2 years ago
- The Truth Is In There: Improving Reasoning in Language Models with Layer-Selective Rank Reduction☆389Updated last year
- ☆558Updated last year
- Finetuning Large Language Models on One Consumer GPU in 2 Bits☆733Updated last year
- ☆551Updated last year
- batched loras☆347Updated 2 years ago
- Fine-tune mistral-7B on 3090s, a100s, h100s☆722Updated 2 years ago
- Memory optimization and training recipes to extrapolate language models' context length to 1 million tokens, with minimal hardware.☆751Updated last year
- Repo for "Monarch Mixer: A Simple Sub-Quadratic GEMM-Based Architecture"☆560Updated 11 months ago
- Generate textbook-quality synthetic LLM pretraining data☆507Updated 2 years ago
- Inference code for Persimmon-8B☆412Updated 2 years ago
- A library for easily merging multiple LLM experts, and efficiently train the merged LLM.☆500Updated last year
- [ICML 2024] SqueezeLLM: Dense-and-Sparse Quantization☆711Updated last year
- A repository for research on medium sized language models.☆524Updated 6 months ago
- Official repository for ORPO☆468Updated last year
- Official code for ReLoRA from the paper Stack More Layers Differently: High-Rank Training Through Low-Rank Updates☆469Updated last year
- Generative Representational Instruction Tuning☆680Updated 5 months ago
- Customizable implementation of the self-instruct paper.☆1,050Updated last year
- Inference code for Mistral and Mixtral hacked up into original Llama implementation☆371Updated 2 years ago
- This repository contains code to quantitatively evaluate instruction-tuned models such as Alpaca and Flan-T5 on held-out tasks.☆551Updated last year