Skip to content

AI/ML benchmark for local LLM inference and XGBoost training on GPU/CPU

AI Coding Tools

A single‑command Python suite that benchmarks GPU/CPU performance on typical AI/ML workloads such as Ollama LLMs and XGBoost, producing an interactive HTML report.

Python Latest v0.6.4 · 3d ago Security brief →

Features

  • One‑command execution of comprehensive GPU/CPU AI/ML benchmarks
  • Supports both Ollama large language model latency/throughput tests and XGBoost training/inference on the HIGGS dataset
  • Generates an automatically opened Jupyter notebook exported as HTML for immediate results
  • Optionally uploads encrypted benchmark data to a central Streamlit dashboard for community comparison

Recent releases

View all 8 releases →
No immediate action
v0.6.2 Feature

Resource detection + flags

No immediate action
v0.5.0 Bugfix

pyarrow → fastparquet; XGBoost CUDA disabled

No immediate action
v0.4.0 Feature

Progress bar + integrity check + Mermaid diagram

No immediate action
v0.3.0 Bugfix

Exception handling improvements

No immediate action
v0.2.0 Feature

Notebook generation + cloud sharing

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

About

Stars
18
Forks
0
Languages
Python Mermaid

Install & Platforms

Install via
brew npm cargo go shell-script binary
Platforms
linux macos windows arm64

Beta — feedback welcome: [email protected]