TareksLab/L2-MERGE4

MERGE4

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

Merge Details

Merge Method

This model was merged using the Linear DELLA merge method using huihui-ai/Llama-3.3-70B-Instruct-abliterated as a base.

Models Merged

The following models were included in the merge:

• ReadyArt/Fallen-Safeword-70B-R1-v4.1
• Sao10K/L3-70B-Euryale-v2.1
• ReadyArt/Fallen-Abomination-70B-R1-v4.1
• TheDrummer/Fallen-Llama-3.3-R1-70B-v1
• ReadyArt/Forgotten-Safeword-70B-3.6

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: Sao10K/L3-70B-Euryale-v2.1
    parameters:
      weight: 0.15
      density: 0.7
      epsilon: 0.2
      lambda: 1.1
  - model: TheDrummer/Fallen-Llama-3.3-R1-70B-v1
    parameters:
      weight: 0.15
      density: 0.7
      epsilon: 0.2
      lambda: 1.1
  - model: ReadyArt/Fallen-Abomination-70B-R1-v4.1
    parameters:
      weight: 0.15
      density: 0.7
      epsilon: 0.2
      lambda: 1.1
  - model: ReadyArt/Forgotten-Safeword-70B-3.6
    parameters:
      weight: 0.15
      density: 0.7
      epsilon: 0.2
      lambda: 1.1
  - model: ReadyArt/Fallen-Safeword-70B-R1-v4.1
    parameters:
      weight: 0.15
      density: 0.7
      epsilon: 0.2
      lambda: 1.1
  - model: huihui-ai/Llama-3.3-70B-Instruct-abliterated
    parameters:
      weight: 0.25
      density: 0.7
      epsilon: 0.1
      lambda: 1.0
base_model: huihui-ai/Llama-3.3-70B-Instruct-abliterated
merge_method: della_linear
parameters:
  normalize: false
tokenizer:
  source: base
dtype: bfloat16