QuantFactory/Llama-Deepsync-3B-GGUF
Full fine-tune vs prithivMLmods/Codepy-Deepthink-3B. QuantFactory/Llama-Deepsync-3B-GGUF is a GGUF (Q8_0) of different weights. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 0/57, MLP 84/84, Attention 112/112