NbAiLab/nb-whisper-small

Full fine-tune vs openai/whisper-small. Fine-tune: 100% of the weights changed, across 12 of 12 layers.

Tensors changed: Embeddings 3/3, Norms 124/124, Attention 252/252, MLP 96/96, other 4/4

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

On the Hub · Downloads on Model Pulse