Hahmdong/SPUPER-qwen3.5-9b-quanquer-add-quanter1
Partial fine-tune vs Hahmdong/SPUPER-qwen3.5-9b-quanquer-add. Partial change: 84% of the weights changed (attention, lm_head, mlp, norms), in 32 of 32 layers.
Tensors changed: Attention 202/332, Norms 93/215, lm_head 1/1, Embeddings 0/4, MLP 96/208