verl
verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework.
HYSEN LABS DIRECTORY
Platform tag · curated repositories and deep analysis.
verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework.
Fully automated version management and package publishing. semantic-release Fully automated version management and package publishing semantic-release** automates the whole package release workflow including: determining the next version number, generating the release notes, and publishing the package.
Project brief: Contexts Optical Compression. Model Download | Paper Link | Arxiv Paper Link | DeepSeek-OCR: Contexts Optical Compression Explore the boundaries of visual-text compression.
Graph Neural Network Library for PyTorch. For this, we load the Cora dataset, and create a simple 2-layer GCN model using the pre-defined GCNConv: We can now optimize the model in a training loop, similar to the standard PyTorch training procedure .
Live reload for Go apps. The legacy build.bin field is deprecated and will be removed in a future release, so prefer the entrypoint form going forward.
The first open-source harness builder for AI coding. Make AI coding deterministic and repeatable.
A reactive notebook for Python : run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.
Wave combines a terminal, editor, browser, and AI assistance in one cross-platform workspace.
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Code for Machine Learning for Trading, 3rd edition, from data sourcing to live execution.
HivisionIDPhotos: a lightweight and efficient AI ID photos tools. AI .
AI Agent Framework, the Pydantic way. Yet despite virtually every Python agent framework and LLM library using Pydantic Validation, when we began to use LLMs in Pydantic Logfire, we couldn't find anything that gave us the same feeling.
SWE-agent takes a GitHub issue and tries to automatically fix it, using your LM of choice. It can also be employed for offensive cybersecurity or competitive coding challenges. [NeurIPS 2024].
Empowering Data Intelligence with Distributed SQL for Sharding, Scalability, and Security Across All Databases.
支持 Python3、JavaScript、Shell、Typescript 的定时任务管理平台(Timed task management platform supporting Python3, JavaScript, Shell, Typescript).
A simpler site generator. Transforms a directory of templates (of varying types) into HTML.
Project brief: High-performance AI pipeline engine with a C++ core and 50+ Python-extensible nodes. Build, debug, and scale LLM workflows with 13+ model providers, 8+ vector databases, and agent orchestration, all from your IDE. Includes VS Code extension, TypeScript/Python SDKs, and Docker deployment.
Train transformer language models with reinforcement learning. Built on top of the Transformers ecosystem, TRL supports a variety of model architectures and modalities, and can be scaled-up across various hardware setups.
ARIS (Auto-Research-In-Sleep), Lightweight Markdown-only skills for autonomous ML research: cross-model review loops, idea discovery, and experiment automation. No framework, no lock-in, works with Claude Code, Codex, OpenClaw, or any LLM agent.
Turns Data and AI algorithms into production-ready web applications in no time.