Gemma3-LoRAtuned-Resta

Gemma3-LoRAtuned-Resta is a merge of the following models using mergekit:

• google/gemma-3-1b-pt
• shubhro2025/gemma-harm

Configuration

```yaml merge_method: task_arithmetic base_model: merged_gemma_lora_full models:

• model: google/gemma-3-1b-pt parameters: weight: 0.2 # α = 0.2
• model: shubhro2025/gemma-harm parameters: weight: -0.2
  ```