jaymanaryan/CleanPool-v3-w10
Full fine-tune vs MihaiPopa-1/Qwen3-0.6B-English-Hinglish-Preview. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 101/113, MLP 84/84, Attention 112/112
Full fine-tune vs MihaiPopa-1/Qwen3-0.6B-English-Hinglish-Preview. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 101/113, MLP 84/84, Attention 112/112