Shiina-Qwen2.5-32B

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

Uncensored Qwen2.5 with a Japanese base model: rinna/qwen2.5-bakeneko-32b

Merge Details

Merge Method

This model was merged using the TIES merge method using rinna/qwen2.5-bakeneko-32b as a base.

Models Merged

The following models were included in the merge:

• nbeerbower/Dumpling-Qwen2.5-32B-v2
• CombinHorizon/zetasepic-abliteratedV2-Qwen2.5-32B-Inst-BaseMerge-TIES
• nbeerbower/Dumpling-Qwen2.5-32B
• nbeerbower/Azura-Qwen2.5-32B
• ArliAI/Qwen2.5-32B-ArliAI-RPMax-v1.3

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: nbeerbower/Azura-Qwen2.5-32B
    parameters:
      weight: 1
      density: 1
  - model: ArliAI/Qwen2.5-32B-ArliAI-RPMax-v1.3
    parameters:
      weight: 1
      density: 1
  - model: CombinHorizon/zetasepic-abliteratedV2-Qwen2.5-32B-Inst-BaseMerge-TIES
    parameters:
      weight: 1
      density: 1
  - model: nbeerbower/Dumpling-Qwen2.5-32B
    parameters:
      weight: 1
      density: 1
  - model: nbeerbower/Dumpling-Qwen2.5-32B-v2
    parameters:
      weight: 1
      density: 1
merge_method: ties
base_model: rinna/qwen2.5-bakeneko-32b
parameters:
  weight: 1
  density: 1
  normalize: true
  int8_mask: true
dtype: bfloat16
tokenizer:
   source: rinna/qwen2.5-bakeneko-32b-instruct