Latest Open Source Projects
deepclaude
deepclaude
TLDR: DeepClaude is a high - performance LLM inference API. It combines DeepSeek R1's reasoning and Anthropic Claude's creative and code generation capabilities, offering features like zero latency, security, high configurability, and is open - source. It allows users to use their own API keys and provides a unified interface to leverage the strengths of both models.
browser-use
browser-use
TLDR: The browser-use repository provides an easy way to connect AI agents with the browser. It offers features like vision and html extraction, multi-tab management, custom actions, and parallelization of agents. It also collects anonymous usage data for improvement.
open-computer-use
open-computer-use
TLDR: A secure cloud Linux computer powered by E2B Desktop Sandbox and controlled by open-source LLMs. Supports various LLMs like Meta Llama and OS-Atlas. Operates via keyboard, mouse and shell commands. Easily add new LLMs adhering to OpenAI API specification.
text-extract-api
text-extract-api
TLDR: A tool for converting images, PDFs, and Office documents to Markdown or JSON with high accuracy. Built with FastAPI, uses Celery for asynchronous tasks and Redis for caching. Supports various OCR strategies and can remove PII. Comes with a CLI tool and has different storage strategies. Also has an online demo and dedicated API clients.
RAG_Techniques
RAG_Techniques
TLDR: This repository is a comprehensive collection of advanced Retrieval-Augmented Generation (RAG) techniques. It includes various methods for enhancing RAG systems such as query enhancement, context enrichment, advanced retrieval methods, iterative and adaptive techniques, evaluation, explainability, and advanced architectures.
Building-llama3-from-scratch
Building-llama3-from-scratch
TLDR: This repository contains code to build the LLaMA 3 language model from scratch using Python. It explains the components of LLaMA 3 such as pre-normalization using RMSNorm, SwiGLU activation function, Rotary Embeddings (RoPE), and Byte Pair Encoding (BPE) Algorithm. The code shows how to tokenize input data, create embeddings for each token, implement attention heads, self-attention, multi-head attention, SwiGLU activation function, and generate the output.
firecrawl
firecrawl
TLDR: Firecrawl is an API service that crawls URLs and converts them into clean markdown or structured data. It offers advanced scraping, crawling, and data extraction capabilities with features like LLM-ready formats, customizability, and more. It also has SDKs for various languages and integrations with multiple frameworks.
FlagEmbedding
FlagEmbedding
TLDR: FlagEmbedding focuses on retrieval-augmented LLMs and consists of multiple projects including inference, finetune, evaluation, dataset, tutorials, and research. It offers various embedding and reranker models for different languages and tasks.
LLMs-from-scratch
LLMs-from-scratch
TLDR: This repository contains code for developing, pretraining, and finetuning a GPT-like LLM. It is the official code repository for the book 'Build a Large Language Model (From Scratch)'. The code is designed to run on conventional laptops and automatically utilizes GPUs if available. It also includes bonus materials and has specific hardware requirements.
llm-course
llm-course
TLDR: This repository contains a comprehensive course on large language models, divided into three parts: LLM Fundamentals, The LLM Scientist, and The LLM Engineer. It includes notebooks and articles related to various aspects of large language models such as fine-tuning, quantization, and building applications.