mfm

Moderation of NSFW media files via AI.

Installation

ONNX Runtime

The project uses ONNX Runtime for ML model inference. Install the ONNX Runtime shared library:

• macOS:

brew install onnxruntime

• Other platforms: See ONNX Runtime installation guide.

Model

Download ONNX NSFW detection models and configure their paths in .env file. The service requires environment variables to be set:

• MODEL_OPEN - absolute path to OpenNSFW2 model
• MODEL_VIT - absolute path to ViT model

Models should accept input shape [batch_size, 224, 224, 3] (float32, normalized 0-1) and output shape [batch_size, 1] (float32, probability of NSFW) for OpenNSFW2 or [batch_size, 2] for ViT (probability pair: normal, nsfw).

If environment variables are not set, the service will panic with an error message.

See details on model conversion: doc/models/<model>.md

ffmpeg

• macOS:

brew install ffmpeg

Other Dependencies

go install github.com/go-task/task/v3/cmd/task@latest
go install github.com/vektra/mockery/v2@v2.53.3
go install github.com/golangci/golangci-lint/v2/cmd/golangci-lint@latest

Environment Variables

Copy the .env.example file to .env and configure the environment variables:

cp .env.example .env

Test content

Create assets/[1..7].png and assets/video[1..2].mp4