lilmeaty / ddfd

merge

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

Merge Details

Merge Method

This model was merged using the passthrough merge method using meta-llama/Llama-3.2-3B-Instruct as a base.

Models Merged

The following models were included in the merge:

• chuanli11/Llama-3.2-3B-Instruct-uncensored
• meta-llama/Llama-3.2-3B
• PJMixers-Dev/LLaMa-3.2-Instruct-JankMix-v0.1-SFT-3B
• stabilityai/stable-code-3b
• ICEPVP8977/Uncensored_llama_3.2_3b_safetensors
• ibm-granite/granite-3b-code-base-2k
• Bllossom/llama-3.2-Korean-Bllossom-3B
• Qwen/Qwen2.5-3B
• ministral/Ministral-3b-instruct
• Qwen/Qwen2.5-3B-Instruct

Configuration

The following YAML configuration was used to produce this model:

slices:
  - sources:
      - layer_range: [0, 2]
        model: ICEPVP8977/Uncensored_llama_3.2_3b_safetensors
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1

  - sources:
      - layer_range: [0, 2]
        model: meta-llama/Llama-3.2-3B-Instruct
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1

  - sources:
      - layer_range: [0, 2]
        model: meta-llama/Llama-3.2-3B
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1

  - sources:
      - layer_range: [0, 2]
        model: chuanli11/Llama-3.2-3B-Instruct-uncensored
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1

  - sources:
      - layer_range: [0, 2]
        model: PJMixers-Dev/LLaMa-3.2-Instruct-JankMix-v0.1-SFT-3B
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1

  - sources:
      - layer_range: [0, 2]
        model: Bllossom/llama-3.2-Korean-Bllossom-3B
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1

  - sources:
      - layer_range: [0, 2]
        model: Qwen/Qwen2.5-3B-Instruct
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1

  - sources:
      - layer_range: [0, 2]
        model: Qwen/Qwen2.5-3B
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1

  - sources:
      - layer_range: [0, 2]
        model: stabilityai/stable-code-3b
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1

  - sources:
      - layer_range: [0, 2]
        model: ibm-granite/granite-3b-code-base-2k
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1

  - sources:
      - layer_range: [0, 2]
        model: ministral/Ministral-3b-instruct
        parameters:
          normalize: true
          int8_mask: true
          density: 0.5
          weight: 1


random_seed: 0
temperature: 0.5
top_p: 0.65
merge_method: passthrough
dtype: float16
base_model: meta-llama/Llama-3.2-3B-Instruct