ONNX Model Training Setup

I’m looking for an expert to help configure a cloud-based server with GPU acceleration (NVIDIA preferred) for custom AI model training in ONNX format. The final goal is to produce optimized ONNX models that can be imported into FaceFusion for face-swapping tasks.

Requirements:

Experience setting up cloud servers (AWS, GCP, or similar) with GPU support

Knowledge of ONNX model training pipelines

Familiarity with FaceFusion or similar face-swap tools

Ability to configure the environment for fast inference and training performance

Optional: experience optimizing models for Apple Silicon / CoreML (bonus)

Deliverables:

Functional cloud server setup with working GPU

Workflow to train or fine-tune custom models in ONNX

Example ONNX model tested and running within FaceFusion

Documentation for how to repeat the process