lokeshkumar79/kid-whisper-tiny-en-myst-ours

Full fine-tune vs openai/whisper-tiny.en. Fine-tune: 98% of the weights changed, across 4 of 4 layers.

Tensors changed: Embeddings 1/3, Norms 44/44, Attention 84/84, MLP 32/32, other 4/4

On the Hub · Downloads on Model Pulse