Etherll/NoisySpeechDetection-v0.2

Full fine-tune (fewer layers) vs unsloth/whisper-small. Etherll/NoisySpeechDetection-v0.2 matches part of unsloth/whisper-small (187 of its 479 tensors; the rest of unsloth/whisper-small has no counterpart). On the shared part: Fine-tune: 99% of the weights changed, across 12 of 12 layers.

Tensors changed: other 4/4, Embeddings 0/1, Norms 39/50, MLP 41/48, Attention 84/84

On the Hub · Downloads on Model Pulse