Skills for Claude Code

C

Claude Monitor

sickn33

Diagnose performance issues with Claude Code and local systems.

Autoresearch Agent

alirezarezvani

Automate your code optimization experiments.

C

CuTile Autotuning

nvidia

Optimize CuTile kernels with automated tuning techniques.

DeepStream Profiling

nvidia

Efficiently profile and tune DeepStream pipelines.

RAG-Perf

nvidia

Benchmark performance of NVIDIA RAG Blueprint servers.

CUDA Graphs

nvidia

Optimize GPU performance with CUDA graph capture.

O

Omniverse USD Performance Tuning

nvidia

Optimize USD performance for better loading and rendering.

J

Jetson Inference Memory Tuning

nvidia

Optimize memory settings for LLM/VLM on NVIDIA Jetson.

Accelerated Computing cuDF

nvidia

Efficient GPU DataFrame operations for pandas users.

C

CPU Offloading

nvidia

Optimize memory usage by offloading CPU tasks in Megatron Bridge.

Activation Recompute

nvidia

Optimize GPU memory usage with activation recompute.

NeMo Evaluator Plugin

nvidia

Evaluate models and datasets efficiently with NeMo.

M

MoE Communication Overlap

nvidia

Optimize expert-parallel communication in Megatron Bridge.

TensorBoard

davila7

Visualize and debug machine learning models effectively.

M

Model Pruning

davila7

Compress LLMs and speed up inference with pruning techniques.

GPTQ Quantization

davila7

Efficiently deploy large LLMs on consumer GPUs with minimal loss.

H

Half-Quadratic Quantization

davila7

Efficient model quantization without calibration data.

GGUF Quantization

davila7

Efficient model deployment for consumer hardware.

Flash Attention Optimization

davila7

Enhance transformer performance with optimized attention.

B

BitsAndBytes Quantization

davila7

Optimize LLMs with efficient quantization techniques.

A

AWQ Quantization

davila7

Optimize large models with 4-bit quantization.

Next.js & React Performance Expert

vudovn

Optimize your React and Next.js applications effectively.

Flash Attention

nousresearch

Accelerate transformer training and inference significantly.

HuggingFace Accelerate

nousresearch

Simplify distributed PyTorch training with minimal code changes.