Build GPU-accelerated Rust ML models with the Candle framework.
npx skills add https://github.com/gar-ai/mallorn --skill rust-candle-core
Please help me install this Agent Skill. Skill: rust-candle-core Source: https://github.com/gar-ai/mallorn/tree/main/.claude/skills/rust-ml-candle-core Command: npx skills add https://github.com/gar-ai/mallorn --skill rust-candle-core
Build native Rust ML models efficiently using the Candle framework, enabling GPU-accelerated pipelines for vision transformers, large language models, and audio processing.
Initialize a Rust project configured with the Candle GPU workflow to begin building a simple Transformer-based model.
High-intent search queries and answers about installing and using this skill.
Build GPU-accelerated Rust ML models using the Candle framework by configuring your Cargo project with candle-core and candle-nn crates, enabling CUDA feature flags for high-performance neural network execution.
Implementing vision transformers in Rust uses Candle patterns like PatchEmbed and multi-head attention, providing structured guidance for loading models and building high-performance GPU pipelines for neural networks.
Candle supports rotary position embeddings (RoPE) and RMSNorm patterns, allowing developers to build and optimize large language models natively in Rust with GPU acceleration.
Using the Candle framework requires a Rust toolchain and Cargo to manage dependencies, while optional CUDA support can be enabled via the candle feature set for GPU acceleration.
Memory-mapped weight loading in Candle enables efficient memory usage by mapping weight files directly into memory, streamlining the deployment of large language models and audio models in Rust.
You can run audio processing models natively in Rust with GPU acceleration using the Candle framework, which provides core abstractions for building high-performance audio neural networks.
dollspace-gayDetect fake GPU-accelerated Rust code and verify actual kernel execution.
RheasilviaOptimize transformer attention with flash-attn for faster long-sequence training.
dojoengineImplement ML/AI workflows in Rust from model loading to inference.
flexisuite-orgGuide ML/AI application development in Rust with Candle, Burn, and polars.
