litagin/anime-whisper

Full fine-tune vs kotoba-tech/kotoba-whisper-v2.0. Fine-tune: 100% of the weights changed, across 32 of 32 layers.

Tensors changed: Embeddings 1/3, Norms 144/144, Attention 252/252, MLP 136/136, other 4/4

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

On the Hub · Downloads on Model Pulse