Qwen/QwQ-32B-GGUF
Full fine-tune vs Qwen/QwQ-32B. Qwen/QwQ-32B-GGUF is a GGUF (FP16) of different weights. Fine-tune: 95% of the weights changed, across 64 of 64 layers.
Tensors changed: Embeddings 0/1, Norms 128/129, MLP 191/192, Attention 256/448, lm_head 0/1