Ready-to-run Dockerfiles for audio processing and ML/DSP workloads. Build once, run anywhere (local CPU/GPU or in a Cloud environment like AWS/GCP).
- Batteries-included images for common audio/ML pipelines with pinned versions for reproducibility..
- CPU/GPU support when available.
- Exact stack per image (Python, CUDA/cuDNN, PyTorch, torchaudio/vision, librosa, NumPy, SciPy, FFmpeg) + OS base.
- Cloud-ready: sample GCP job specs to deploy at scale.
- CPU/GPU variants with matching CUDA toolkits.
# Build a tool/image
docker build -t audio-tools:latest .