Avacyn/qwen3-1.7B-french-instruct-GGUF
Full fine-tune vs unsloth/Qwen3-1.7B. Avacyn/qwen3-1.7B-french-instruct-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