romange

Roman Gershman's starred repositories

RWKV-LM

RWKV is an RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). So it's combining the best of RNN and transformer - great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding.

Language:PythonApache-2.012418 133 204

ann-benchmarks

Benchmarks of approximate nearest neighbor libraries in Python

Language:PythonMIT4867 117 207

llmware

Unified framework for building enterprise RAG pipelines with small, specialized models

Language:PythonApache-2.04525 43 131

matmulfreellm

Implementation for MatMul-free LM.

Language:PythonApache-2.02881 43 29

cpu_features

A cross platform C99 library to get cpu features at runtime.

Language:C++Apache-2.02437 1202 117

sdsl-lite

Succinct Data Structure Library 2.0

Language:C++NOASSERTION2205 119 199

piko

An open-source alternative to Ngrok, designed to serve production traffic and be simple to host (particularly on Kubernetes)

Language:GoMIT1840 8 12

jvector

JVector: the most advanced embedded vector search engine

Language:JavaApache-2.01466 30 131

restate

Restate is the platform for building resilient applications that tolerate all infrastructure faults w/o the need for a PhD.

Language:RustNOASSERTION1445 11 819

DiskANN

Graph-structured Indices for Scalable, Fast, Fresh and Filtered Approximate Nearest Neighbor Search

Language:C++NOASSERTION1066 26 195

dragonbox

Reference implementation of Dragonbox in C++

Language:C++Apache-2.0600 21 44

VectorDBBench

A Benchmark Tool for VectorDB

Language:PythonMIT507 15 104

c-blosc2

A fast, compressed, persistent binary data store library for C.

Language:CNOASSERTION438 24 194

libdivsufsort

A lightweight suffix-sorting library

Language:CMIT362 31 22

addr2line

A cross-platform `addr2line` clone written in Rust, using `gimli`

Language:RustApache-2.0298 9 92

fakeredis-py

Implementation of Redis in python without having a Redis server running. Fully compatible with using redis-py.

Language:PythonBSD-3-Clause281 4 159

vector-db-benchmark

Framework for benchmarking vector search engines

Language:PythonApache-2.0266 7 58

suffix-trees

Python implementation of Suffix Trees and Generalized Suffix Trees.

Language:PythonMIT122 6 13

sdsl

Succinct Data Structure Library

Language:C++NOASSERTION105 19 18

vector-cookbook

Timescale Vector Cookbook. A collection of recipes to build applications with LLMs using PostgreSQL and Timescale Vector.

Language:Jupyter Notebook100 10 4

Grisu-Exact

Reference implementation of Grisu-Exact in C++

Language:C++Apache-2.058 7 3

puffinn

Parameterless and Universal FInding of Nearest Neighbors

Language:C++MIT56 8 7

prometheus-redistimeseries-adapter

Prometheus remote storage adapter for RedisTimeSeries

Language:GoBSD-3-Clause33 7 9

gsufsort

gsufsort: building suffix arrays, LCP-arrays and BWTs for string collections [AMB 2020]

Language:CGPL-3.025 8 4

gsa-is

Inducing enhanced suffix arrays for string collections [DCC'16, TCS 2017]

Language:CMIT23 4 2

suffix-tree

A C++ Implementation for Generalized Suffix Trees

Language:C++21 2 3

tinyhnsw

build your own vector database -- the littlest hnsw

Language:Python19 2 2

Penjing

C++20 Generalized Suffix Tree implementation

Language:C++MIT11 2 4

rag-with-amazon-bedrock-and-memorydb

Question Answering Generative AI application with Large Language Models (LLMs), Amazon Bedrock, and Amazon MemoryDB for Redis

Language:Jupyter NotebookMIT-07 20

Algorithms-on-Strings

Trie, Suffix Tree, Suffix Array, Burrows–Wheeler Transform, Shortest Non-Shared Substring, BW Matching, Pattern Matching with Suffix Array, Suffix Tree from Suffix Array.

Language:C++MIT2 20