as1605/Deepfake-audio-detection-V2

Full fine-tune vs mo-thecreator/Deepfake-audio-detection. Fine-tune: 100% of the weights changed, across 12 of 12 layers.

Tensors changed: other 13/13, Norms 54/54, Attention 96/96, MLP 48/48, Embeddings 4/4

2 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse