Mungert/Holo1-3B-GGUF
Full fine-tune (fewer layers) vs Qwen/Qwen2.5-VL-3B-Instruct. Mungert/Holo1-3B-GGUF matches part of Qwen/Qwen2.5-VL-3B-Instruct (434 of its 824 tensors; the rest of Qwen/Qwen2.5-VL-3B-Instruct has no counterpart). On the shared part: Mungert/Holo1-3B-GGUF is a GGUF (Q8_0) of different weights. Fine-tune: 100% of the weights changed, across 36 of 36 layers.
Tensors changed: Embeddings 1/1, Norms 73/73, MLP 108/108, Attention 252/252