puneetkakkar/Bitnet-1.58B
Bitnet 1.58b: This project implements the innovative 1-bit LLM architecture described in recent whitepapers, focusing on efficient training, inference, and open-source collaboration.
This project implements a highly efficient architecture for large language models (LLMs) that uses only 1-bit or 1.58-bit representations, significantly reducing memory and computational demands. It helps AI researchers and machine learning engineers by providing an open-source framework to train and deploy these compact LLMs. You can input standard training datasets and get out smaller, faster, and more energy-efficient language models.
No commits in the last 6 months.
Use this if you are an AI researcher or machine learning engineer looking to experiment with or deploy extremely memory-efficient and computationally light large language models.
Not ideal if you are a general user simply looking to run an existing LLM without needing to understand or optimize its underlying architecture.
Stars
9
Forks
—
Language
Python
License
—
Category
Last pushed
Jun 14, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/puneetkakkar/Bitnet-1.58B"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
huawei-csl/SINQ
Welcome to the official repository of SINQ! A novel, fast and high-quality quantization method...
SILX-LABS/QUASAR-SUBNET
QUASAR is a long-context foundation model and decentralized evaluation subnet built on Bittensor,
stackblogger/bitnet.js
BitNet.Js - A node.js implementation of the microsoft bitnet.cpp inference framework.
m96-chan/0xBitNet
Run BitNet b1.58 ternary LLMs with WebGPU — in browsers and native apps
AnswerDotAI/cold-compress
Cold Compress is a hackable, lightweight, and open-source toolkit for creating and benchmarking...