# Union Tokenizer Generation
base_model: mistralai/Mistral-7B-v0.1
models:
  - model: mistralai/Mistral-7B-Instruct-v0.2
    parameters:
      weight: 1.0
  - model: TIGER-Lab/MAmmoTH2-7B
    parameters:
      weight: 1.0
  - model: Nondzu/Mistral-7B-codealpaca-lora
    parameters:
      weight: 1.0
merge_method: task_arithmetic
parameters:
  normalize: false
  lambda: 1.0
dtype: float16
tokenizer:
  source: union

lejelly/union_tokenizer-mistral-7b-instrcut-math-code

union_tokenizer

Merge Details

Merge Method

Models Merged

Configuration

lejelly/
union_tokenizer-mistral-7b-instrcut-math-code