Search Results for "dataset"

...Apprentice agents can be paired with mentor agents, human reviewers, or domain experts depending on the selected mode. The project supports Codex, Cursor, Claude Code, OpenClaw, OpenCode, Hermes Agent, and custom agent commands. Its seed dataset includes curated tasks, reusable lessons, execution traces, work episodes, and structured experience records. It is designed for long-horizon agent work, workflow evaluation, runtime training, and shared improvement across an agent ecosystem.

OpenHarness is an open-source framework developed to support large-scale machine learning workflows, particularly in the context of training, evaluating, and benchmarking AI models. It provides a structured environment for orchestrating experiments, managing datasets, and standardizing evaluation processes across different models. The project focuses on reproducibility and scalability, allowing researchers and engineers to run consistent experiments while tracking results effectively. It...

CUDA Agent is a research-driven agentic reinforcement learning system designed to automatically generate and optimize high-performance CUDA kernels for GPU workloads. The project addresses the long-standing challenge that efficient CUDA programming typically requires deep hardware expertise by training an autonomous coding agent capable of iterative improvement through execution feedback. Its architecture combines large-scale data synthesis, a skill-augmented CUDA development environment,...

...It reliably identifies interactable icons within user interfaces and understands the semantics of various elements in a screenshot, associating intended actions with the correct screen regions. To achieve this, OmniParser curates an interactable icon detection dataset containing 67,000 unique screenshot images labeled with bounding boxes of interactable icons derived from DOM trees. Additionally, a collection of 7,000 icon-description pairs is used to fine-tune a caption model that extracts the functional semantics of detected elements. Evaluations on benchmarks such as SeeClick, Mind2Web, and AITW demonstrate that OmniParser outperforms GPT-4V baselines, even when using only screenshot inputs without additional information.