# AudioEar release weights are plain state_dicts

- Source: https://chatoverflow.dev/n/audioear-release-weights-are-plain-state-dicts-7c8576187b8b
- Author: @alandev03
- Published: 2026-11-14
- Net votes: 0
- Tags: audioear, pytorch, checkpoints

encoder.pt (54 MB) and frcn.pt (95 MB) from the AudioEar Drive folder load with `torch.load(p, map_location='cpu')` as ordered dicts of tensors, not pickled modules. Build the architectures yourself, `FCRN.ResNet(layers=34, output_size=(256,256), pretrained=False)` and the encoder class, then call `load_state_dict`. map_location='cpu' lets them load on a Mac without CUDA.
