Search Results for "block"
Sort By:
Showing 58 open source projects for "block"
View related business solutions-
$300 Free Credits to Build on Google CloudPut your $300 in credit toward real workloads, then keep building with free monthly usage for 20+ products. No commitment and no charge until you upgrade.
-
Earn up to 16% annual interest with Nexo.Generate interest, borrow against your crypto, and trade a range of cryptocurrencies — all in one platform. Geographic restrictions, eligibility, and terms apply.
-
1
jev-trader
One AI trade decision every Monad block. Jev on Kuru MON-USDC
...Server endpoints expose snapshots, history, fills, positions, costs, and block events through Server-Sent Events. A web dashboard can visualize the strategy and its ongoing results. -
2
DFlash
Block Diffusion for Ultra-Fast Speculative Decoding
DFlash is an open-source framework for ultra-fast speculative decoding using a lightweight block diffusion model to draft text in parallel with a target large language model, dramatically improving inference speed without sacrificing generation quality. It acts as a “drafter” that proposes likely continuations which the main model then verifies, enabling significant throughput gains compared to traditional autoregressive decoding methods that generate token by token. -
3
MoBA
MoBA: Mixture of Block Attention for Long-Context LLMs
MoBA, short for Mixture of Block Attention, is an open-source research implementation of a novel attention mechanism designed to improve the efficiency of large language models processing extremely long contexts. The architecture adapts ideas from Mixture-of-Experts networks and applies them directly to the attention mechanism of transformer models. Instead of forcing each token to attend to every other token in the sequence, MoBA divides the context into blocks and dynamically routes queries to only the most relevant segments of information. ... -
4
LongCat-Video
Foundational video generation model with 13.6B parameters
...It is pretrained for continuation, allowing it to create minutes-long sequences while limiting color drift and quality loss. A coarse-to-fine strategy operates across time and space to produce 720p video at 30 frames per second efficiently. Block Sparse Attention reduces high-resolution inference costs, while multi-reward GRPO training improves visual quality and prompt alignment. The repository includes inference scripts for single- and multi-GPU execution, model-download instructions, and interactive generation examples. It also provides audio-driven Avatar variants for expressive single- or multi-character animation. -
Go from Code to Production URL in SecondsSkip the Kubernetes configs. Cloud Run handles HTTPS, scaling, and infrastructure automatically. Two million requests free per month.
-
5
The Web MCP
A powerful Model Context Protocol (MCP) server
...Ecosystem materials and examples show how it plugs into MCP-capable runtimes and agent frameworks. Overall, the project is aimed at making web intelligence a reliable building block for agent workflows. -
6
camofox-browser
Headless browser automation server for AI agents to visit sites
camofox-browser is a headless browser automation server built specifically for AI agents that need to interact with websites that often block standard automation stacks. It wraps Camoufox, a Firefox fork that performs fingerprint spoofing at the C++ level, which means many browser characteristics are altered before page scripts can inspect them, rather than relying on JavaScript-layer stealth patches. The project is designed around a REST API, making it easier for agents and external tools to create tabs, navigate pages, click elements, type input, scroll, capture screenshots, and manage browsing sessions programmatically. ... -
7
HASH
The best way to use and work with blocks
This is HASH's public monorepo which contains our public code, docs, and other key resources. HASH is a platform for decision-making, which helps you integrate, understand and use data in a variety of different ways. HASH does this by combining various different powerful tools together into one simple interface. These range from data pipelines and a graph database, through to an all-in-one workspace, no-code tool builder, and agent-based simulation engine. These exist at varying stages of... -
8
FlashMLA
FlashMLA: Efficient Multi-head Latent Attention Kernels
...It provides optimized kernels for MLA decoding, including support for variable-length sequences, helping reduce latency and increase throughput in model inference systems using that attention style. The library supports both BF16 and FP16 data types, and includes a paged KV cache implementation with a block size of 64 to efficiently manage memory during decoding. On very compute-bound settings, it can reach up to ~660 TFLOPS on H800 SXM5 hardware, while in memory-bound configurations it can push memory throughput to ~3000 GB/s. The team regularly updates it with performance improvements; for example, a 2025 update claims 5 % to 15 % gains on compute-bound workloads while maintaining API compatibility. -
9
OpenMythos
A theoretical reconstruction of the Claude Mythos architecture
...The project explores the idea that instead of stacking hundreds of unique transformer layers, a smaller set of layers can be reused iteratively during inference to achieve deeper reasoning without increasing parameter count. It divides computation into three main stages, including a pre-processing phase, a looped recurrent reasoning block, and a final output refinement stage, creating a structured pipeline for inference. The architecture incorporates advanced techniques such as mixture-of-experts routing, adaptive computation time, and multiple attention mechanisms to dynamically allocate compute where needed. It is highly configurable through a centralized configuration system, allowing experimentation with different architectural parameters such as loop depth, attention type. -
Train ML Models With SQL You Already KnowBuild and deploy ML models using familiar SQL. Automate data prep with built-in Gemini. Query 1 TB and store 10 GB free monthly.
-
10
Eidos
An extensible framework for Personal Data Management
Eidos is an extensible personal data management platform designed to help users organize and interact with their information using a local-first architecture. The system transforms SQLite into a flexible personal database that can store structured and unstructured information such as notes, documents, datasets, and knowledge resources. Its interface is inspired by tools like Notion, allowing users to create documents, databases, and custom views to organize personal information. Unlike... -
11
dtreeviz
Python library for decision tree visualization & model interpretation
A python library for decision tree visualization and model interpretation. Decision trees are the fundamental building block of gradient boosting machines and Random Forests(tm), probably the two most popular machine learning models for structured data. Visualizing decision trees is a tremendous aid when learning how these models work and when interpreting models. The visualizations are inspired by an educational animation by R2D3; A visual introduction to machine learning. -
12
AutoGPT
Powerful tool that lets you create and run intelligent agents
AutoGPT is an experimental open-source application showcasing the capabilities of the GPT-4 language model. This program, driven by GPT-4, chains together LLM "thoughts", to autonomously achieve whatever goal you set. As one of the first examples of GPT-4 running fully autonomously, AutoGPT pushes the boundaries of what is possible with AI. -
13
FLEXible
Federated Learning (FL) experiment simulation in Python
FLEXible (Federated Learning Experiments) is a Python framework offering tools to simulate FL with deep learning. It includes built-in datasets (MNIST, CIFAR10, Shakespeare), supports TensorFlow/PyTorch, and has extensions for adversarial attacks, anomaly detection, and decision trees. -
14
Venom
Venom is the most complete javascript library for Whatsapp
Venom is a high-performance system developed with JavaScript to create a bot for WhatsApp, support for creating any interaction, such as customer service, media sending, sentence recognition based on artificial intelligence and all types of design architecture for WhatsApp. It's a high-performance alternative API to whatzapp, you can send, text messages, files, images, videos and more. Remember, the API was developed on a platform called RESTful Web services, providing interoperability... -
15
GitHub Copilot for Xcode
AI coding assistant for Xcode
Copilot for Xcode brings GitHub Copilot’s AI-driven code suggestions directly into Apple’s Xcode IDE, giving iOS and macOS developers predictive completions, context-aware recommendations, and natural language assistance while they write Swift, Objective-C, and related code. It embeds seamlessly into the Xcode editor UI, offering completions as you type, including full lines or blocks of code derived from surrounding context and doc comments. Because the integration understands the structure... -
16
Superagent
Superagent protects your AI applications
Superagent is an open-source AI safety platform built to protect applications from prompt injections, data leaks, and harmful outputs. It embeds real-time safety directly into AI workflows, helping teams secure models before threats cause damage. Superagent provides guardrails that block jailbreaks, prompt manipulation, and sensitive data exfiltration. It includes redaction tools to remove PII, PHI, and secrets automatically from text. The platform also scans code repositories to detect AI-specific attack vectors like repo poisoning. Superagent is designed for low-latency production environments and works with any major LLM provider. ... -
17
CUTLASS
CUDA Templates for Linear Algebra Subroutines
...It incorporates strategies for hierarchical decomposition and data movement similar to those used to implement cuBLAS and cuDNN. CUTLASS decomposes these "moving parts" into reusable, modular software components abstracted by C++ template classes. These thread-wide, warp-wide, block-wide, and device-wide primitives can be specialized and tuned via custom tiling sizes, data types, and other algorithmic policy. The resulting flexibility simplifies their use as building blocks within custom kernels and applications. To support a wide variety of applications, CUTLASS provides extensive support for mixed-precision computations, providing specialized data-movement and multiply-accumulate abstractions for half-precision floating point (FP16), BFloat16 (BF16), Tensor Float 32 (TF32), etc. -
18
Corsair
Your Agent's Integration Layer
Corsair is a unified integration layer that lets AI agents interact with external applications without exposing the underlying credentials to the agent. Developers connect integrations once and expose typed tools through Corsair or its MCP interface. A permission system can allow, require approval for, or block read, write, and destructive operations. Credentials are resolved internally and can be stored with envelope encryption rather than passed into model context. Multi-tenant mode isolates each tenant's credentials, data, and permission evaluation. Plugins can also provide signature-verified webhooks through a shared endpoint. Corsair is designed for production agent systems that need OAuth, scoped authorization, human approval, and centralized integration control. -
19
Attention Residuals (AttnRes)
Drop-in replacement for standard residual connections in Transformers
Attention Residuals is a research-driven architectural innovation for transformer-based models that replaces traditional residual connections with an attention-based mechanism to improve information flow across layers. In standard transformers, residual connections simply sum outputs from previous layers, which can lead to uncontrolled growth of hidden states and dilution of early-layer information in deep networks. Attention Residuals introduces a learnable softmax attention mechanism that... -
20
mcp-server-chatsum
Query and Summarize your chat messages
mcp-server-chatsum is an MCP server that indexes your chat history and provides tools to query and produce focused summaries on demand. It offers a simple flow: point the server at a local chat database, run the companion chatbot to ingest messages, and then use the MCP tool to retrieve scoped threads and generate concise syntheses. The tool design lets agents filter by participants, time ranges, or keywords before summarizing, which keeps outputs relevant and reduces hallucinated context.... -
21
AutoKeras
AutoML library for deep learning
...AutoKeras supports several tasks with extremely simple interface. AutoKeras would search for the best detailed configuration for you. Moreover, you can override the base classes to create your own block. -
22
Unla
Gateway service that instantly transforms existing MCP Servers
...Because it is itself MCP-speaking, Unla can sit in front of mixed fleets and normalize transports and schemas for clients. Documentation and pkg.go.dev pages reinforce the positioning as a stable, Go-native building block for MCP deployments. -
23
Harmonist
Portable AI agent orchestration with mechanical protocol enforcement
...The framework includes a catalog of specialized agents, validated memory behavior, supply-chain checks, and hooks that gate code-changing turns. If required reviewers do not run, memory is not updated, or shipped files fail integrity checks, Harmonist can block the workflow from completing. The project uses Python, has no runtime dependencies beyond the standard library, and is positioned as a drop-in agent coordination pack. Its purpose is to bring structure, review discipline, and repeatable process control to AI-assisted development. -
24
Hugging Face Skills
Definitions for AI/ML tasks like dataset creation
...The project is designed to be interoperable across major agent ecosystems, including Claude Code, OpenAI Codex, Gemini CLI, and Cursor, making it a cross-platform building block for agent automation. By formalizing best practices and workflows, Skills helps transform general-purpose coding agents into domain-aware assistants that can execute complex ML pipelines with less manual prompting. The repository also includes ready-to-use skills for common Hugging Face operations and encourages teams to extend them with custom domain logic. -
25
NextJS Ollama LLM UI
Fully-featured web interface for Ollama LLMs
...The interface stores conversations in local storage, so no separate backend database is required, making it ideal for hobbyists, experimenters, and developers who want a simple, web-accessible portal to their models. It includes usability enhancements like code syntax highlighting and easy code block copying, plus basic controls to download and manage models directly from the web UI.
