dbaysal/metallama3.2-3b-learned

Adapter vs meta-llama/Llama-3.2-3B. dbaysal/metallama3.2-3b-learned is a LoRA adapter, not full weights: rank 64, on down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj in 28 layers (196 adapted matrices). It modifies meta-llama/Llama-3.2-3B at load time.

On the Hub · Downloads on Model Pulse