models:
  - model: meta-llama/Llama-2-7b-hf
    parameters:
      weight: 0.5
  - model: epfl-llm/meditron-7b
    parameters:
      weight: 0.5

merge_method: task_arithmetic
base_model: meta-llama/Llama-2-7b-hf
parameters:
  normalize: true
  int8_mask: true

dtype: bfloat16

Chayaaaaa/without_bio_apart_from_Llama-2-7B-hf

without_bio_apart_from_Llama-2-7B-hf

🧩 Configuration

Chayaaaaa/
without_bio_apart_from_Llama-2-7B-hf