deepdml/whisper-base-ar-quran-mix-norm
Full fine-tune vs openai/whisper-base. Fine-tune: 99% of the weights changed, across 6 of 6 layers.
Tensors changed: Embeddings 1/3, Norms 64/64, Attention 126/126, MLP 48/48, other 4/4
Full fine-tune vs openai/whisper-base. Fine-tune: 99% of the weights changed, across 6 of 6 layers.
Tensors changed: Embeddings 1/3, Norms 64/64, Attention 126/126, MLP 48/48, other 4/4