primeline/whisper-large-v3-turbo-german
Full fine-tune (fewer layers) vs primeline/whisper-large-v3-german. primeline/whisper-large-v3-turbo-german matches part of primeline/whisper-large-v3-german (587 of its 1259 tensors; the rest of primeline/whisper-large-v3-german has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Embeddings 2/3, Norms 156/156, Attention 280/280, MLP 144/144, other 4/4
2 other repos hold byte-for-byte identical weight files.