jeiku/Smaura_8B
Adapter vs ResplendentAI/Aura_Llama3. ResplendentAI/Aura_Llama3 is a LoRA adapter, not full weights: rank 128, on q_proj, v_proj in 32 layers (64 adapted matrices). It modifies jeiku/Smaura_8B at load time.
2 derivatives checked
- mradermacher/Smaura_8B-i1-GGUF: Quantization
- mradermacher/Smaura_8B-GGUF: Same weights, converted