mlabonne/OrpoLlama-3-8B
LoRA-style vs meta-llama/Meta-Llama-3-8B. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: Norms 0/65, MLP 96/96, Attention 128/128
5 derivatives checked
- QuantFactory/OrpoLlama-3-8B-GGUF: Full fine-tune
- mradermacher/OrpoLlama-3-8B-i1-GGUF: Quantization
- mradermacher/OrpoLlama-3-8B-GGUF: Same weights, converted
- IsakNordgren/NeuralDaredevil-8B-abliterated-x-OrpoLlama-3-8B: Full fine-tune
- IsakNordgren/Llama-3-8B-instruct-x-OrpoLlama-3-8B: Full fine-tune