vxef/smol_lora_toy
Adapter vs HuggingFaceTB/SmolLM-135M. vxef/smol_lora_toy is a LoRA adapter, not full weights: rank 8, on q_proj, v_proj in 30 layers (60 adapted matrices). It modifies HuggingFaceTB/SmolLM-135M at load time.
Adapter vs HuggingFaceTB/SmolLM-135M. vxef/smol_lora_toy is a LoRA adapter, not full weights: rank 8, on q_proj, v_proj in 30 layers (60 adapted matrices). It modifies HuggingFaceTB/SmolLM-135M at load time.