caffe-cifar-10

Build Caffe from source and train CNNs on CIFAR-10.

Updated Apr 14, 2026
One-click install
npx skills add https://github.com/bianhaifeng789-hue/openclaw-config --skill caffe-cifar-10-bianhaifeng789-hue
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: caffe-cifar-10
Source: https://github.com/bianhaifeng789-hue/openclaw-config/tree/main/skills/tb2/caffe-cifar-10
Command: npx skills add https://github.com/bianhaifeng789-hue/openclaw-config --skill caffe-cifar-10-bianhaifeng789-hue

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides practical, step-by-step guidance to compile the legacy Caffe framework from source and train convolutional models on the CIFAR-10 dataset while addressing common compatibility issues on modern systems.

Core Features & Use Cases

  • Source build guidance: Detailed workflow for cloning the correct Caffe release, installing dependencies, configuring Makefile.config, and applying source patches.
  • Compatibility handling: Advice for OpenCV 4.x API changes, Python header and numpy path issues, compiler/CUDA considerations, and conservative build parallelism to avoid OOM.
  • Training workflow: Instructions for preparing CIFAR-10 data, adjusting solver/prototxt settings, running training and tests, and verifying LMDB/LevelDB outputs and model snapshots.
  • Use case: Reproducing and training the Caffe CIFAR-10 example on a modern Ubuntu system where OpenCV and Python versions differ from legacy expectations.

Quick Start

Follow the checklist to verify dependencies, patch OpenCV/Python incompatibilities, build Caffe conservatively, run make runtest, prepare the CIFAR-10 dataset, and launch the example training script.

Frequently Asked Questions about caffe-cifar-10

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build Caffe from source on a modern Ubuntu system with OpenCV 4?

To build Caffe from source, clone the correct release, install dependencies, configure Makefile.config, and apply OpenCV 4 compatibility patches to address API changes before compiling conservatively to avoid out-of-memory errors.

What is the correct way to prepare the CIFAR-10 dataset for Caffe training?

Preparing the CIFAR-10 dataset for Caffe training involves running conversion scripts to generate and verify LMDB or LevelDB outputs, ensuring the data is correctly formatted before adjusting solver and prototxt settings.

Why does my Caffe compilation fail with Python or numpy path issues?

Caffe compilation fails with Python or numpy path issues due to misconfigured Makefile.config settings, requiring you to verify dependency paths, correctly set Python headers and numpy paths, and manage compiler compatibility.

Can I use conservative make parallelism to avoid OOM when building Caffe?

Yes, you can and should use conservative make parallelism when building Caffe to avoid out-of-memory errors, especially when managing compiler and CUDA compatibility issues during the source compilation process.

How do I verify a successful Caffe installation after compiling from source?

To verify a successful Caffe installation after compiling from source, run the post-build verification command make runtest to check build integrity and ensure all dependencies and configurations are correctly applied.