classla/Wav2Vec2BertPrimaryStressAudioFrameClassifier

Full fine-tune vs facebook/w2v-bert-2.0. Fine-tune: 100% of the weights changed, across 24 of 24 layers.

Tensors changed: other 74/74, Norms 290/290, MLP 192/192, Embeddings 24/24, Attention 192/192

On the Hub · Downloads on Model Pulse