Guilherme34/Firefly
Full fine-tune vs DavidAU/Llama-3.2-3B-Instruct-heretic-ablitered-uncensored. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 10/57, MLP 84/84, Attention 112/112
3 derivatives checked
- mradermacher/Firefly-i1-GGUF: Quantization
- mradermacher/Firefly-GGUF: Same weights, converted
- Guilherme34/Firefly-V2: Full fine-tune