Hjgugugjhuhjggg/mergekit-dare_linear-caalfar

merge

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the linear DARE merge method using meta-llama/Llama-3.2-1B as a base.

Models Merged

The following models were included in the merge:

• autoprogrammer/Llama-3.2-1B-Instruct-MGSM8K-sft1
• autoprogrammer/Llama-3.2-1B-Instruct-medmcqa-zh-linear
• Alelcv27/llama3.2-1b-math-code
• huyhoangt2201/llama-3.2-1b-sql_finetuned_billingual_3.0_merged
• meta-llama/Llama-3.2-1B-Instruct

Configuration

The following YAML configuration was used to produce this model:

merge_method: dare_linear
architectures: ["transformer"]
base_model: meta-llama/Llama-3.2-1B

models:
  - model: Alelcv27/llama3.2-1b-math-code
  - model: huyhoangt2201/llama-3.2-1b-sql_finetuned_billingual_3.0_merged
  - model: autoprogrammer/Llama-3.2-1B-Instruct-MGSM8K-sft1
  - model: meta-llama/Llama-3.2-1B-Instruct
  - model: autoprogrammer/Llama-3.2-1B-Instruct-medmcqa-zh-linear

parameters:
  density: 0.5
  weight: 1.0
  int8_mask: true