bytkim/Qwen3.6-27B-MTP-pi-reasoning-GGUF
Full fine-tune vs Qwen/Qwen3.6-27B. bytkim/Qwen3.6-27B-MTP-pi-reasoning-GGUF is a GGUF (BF16) of different weights. Fine-tune: 100% of the weights changed, across 64 of 64 layers.
Tensors changed: Embeddings 1/1, MLP 192/192, Attention 64/64, lm_head 1/1, Norms 0/161