alkhrzmy/qwen3-tts-0.6b-indonesian
Partial fine-tune vs Qwen/Qwen3-TTS-12Hz-0.6B-Base. Partial change: 65% of the weights changed (attention, embeddings, lm_head, mlp, norms, other), in 28 of 28 layers.
Tensors changed: lm_head 14/15, Embeddings 15/17, Norms 4/134, MLP 103/103, Attention 132/132, other 1/1
1 other repos hold byte-for-byte identical weight files.