Professor/kinyarwanda-tts-0.6b-full-finetuning
Full fine-tune vs OuteAI/OuteTTS-1.0-0.6B. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 72/113, MLP 84/84, Attention 112/112
Full fine-tune vs OuteAI/OuteTTS-1.0-0.6B. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 72/113, MLP 84/84, Attention 112/112