A high-performance CLI toolset for managing tensors and building neural networks with native Vulkan GPU support.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install rasptorch-cli
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install rasptorch-cli using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Rasptorch?
Rasptorch is a powerful AI development tool designed to bridge the gap between low-level tensor operations and high-level model building. As part of the ecosystem of Openclaw Skills, it provides developers with an agent-native interface to manage machine learning workflows directly from the command line. Whether you are running on Windows or Linux, Rasptorch ensures optimal performance by intelligently switching between CPU and GPU compute via its Vulkan integration.
The skill is built for automation, prioritizing JSON outputs that allow AI agents to parse and react to machine learning state changes in real-time. From simple tensor manipulation to complex Transformer and CNN architectures, Rasptorch offers a streamlined path for rapid prototyping and model training without the overhead of heavy IDEs.
Rasptorch Use Cases
- Automating the creation and inspection of tensors using random, zeros, or ones initialization.
- Verifying hardware acceleration status and Vulkan GPU availability on host machines.
- Building and prototyping various neural network architectures like MLPs, CNNs, GRUs, and Transformers.
- Executing stateful model training sessions with configurable hyperparameters like learning rate and batch size.
- Saving and loading model weights for persistent machine learning experiments within Openclaw Skills environments.
How Rasptorch Works
- The Openclaw agent invokes the rasptorch CLI using either the global console script or the python module invocation.
- The system performs a hardware check to determine if Vulkan-supported GPUs are available for computation.
- The user or agent specifies a compute device using the --device flag (auto, gpu, or cpu).
- Tensors or models are initialized in memory and assigned unique identifiers within a session-like registry.
- Commands are executed with the --json flag to ensure the output is returned in a structured, machine-readable format.
- For stateful operations, the model registry tracks active architectures for training, evaluation, or disk serialization.
Rasptorch Setup
To get started with this entry in the Openclaw Skills library, ensure you have python and the rasptorch binary available in your path.
# Verify installation and GPU status
rasptorch --json info
# Alternative invocation if script is not in PATH
python -m rasptorch.CLI.cli --json info
Rasptorch Data Schema & Taxonomy
Rasptorch utilizes a structured JSON schema for all outputs and handles persistent data through standard file formats.
| Data Type | format | Description |
|---|---|---|
| Command Output | JSON | Structured data including vulkan_available, vulkan_status, and tensor metadata. |
| Model Storage | .pkl | Python pickle files used for saving and loading model weights and architectures. |
| Model Registry | In-Memory | A session-based registry mapping model-id to active neural network instances. |
Rasptorch Advanced Features
- Multi-architecture support including Linear, MLP, CNN, GRU, and Transformer models.
- Automatic hardware fallback logic that prioritizes GPU compute but maintains CPU stability.
- Integrated Streamlit UI support for visual monitoring of machine learning workflows.
- Interactive REPL mode via the chat command for direct human-in-the-loop experimentation.
- Extensive optimizer support including Adam and other standard training algorithms.
SKILL.md
Loading
Related Openclaw Skills
Text to Video Deevid
Generate professional 1080p AI videos from 150-word scripts or simple text prompts in minutes using Openclaw Skills.
php-sql-fixer
A specialized security tool for auditing PHP/Yaf projects to identify SQL injection vulnerabilities and generate safe, parameterized query patches.
Horizontal-Vertical Analysis Framework
A sophisticated dual-axis research methodology that synthesizes historical evolution with contemporary competitive landscapes to uncover hidden strategic insights.
Video Generator Free No Limit
A professional AI video creation tool that transforms text and images into 1080p MP4 videos through a cloud rendering pipeline without usage caps.
Free Video Generator by Script
A powerful AI skill that converts written scripts into high-quality 1080p MP4 videos using remote GPU rendering.
Free Video Generator Like Veo
A sophisticated AI skill that automates cinematic video generation from text prompts using remote GPU rendering.
Featured*
