ChaosFlowerRP-24B

ChaosFlowerRP-24B is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the TIES merge method using trashpanda-org/MS-24B-Instruct-Mullein-v0 as a base.

Models Merged

The following models were included in the merge:

• h34v7/DansXPantheon-RP-Engine-V1.2-24b-Small-Instruct-Ties-Merge
• OddTheGreat/Apparatus_24B

Configuration

The following YAML configuration was used to produce this model:

slices:
  - sources:
      - model: trashpanda-org/MS-24B-Instruct-Mullein-v0
        layer_range: [0, 40]
        parameters:
          density: 0.8
          weight: 0.4
      - model: h34v7/DansXPantheon-RP-Engine-V1.2-24b-Small-Instruct-Ties-Merge
        layer_range: [0, 40]
        parameters:
          density: 0.7
          weight: 0.3
      - model: OddTheGreat/Apparatus_24B
        layer_range: [0, 40]
        parameters:
          density: 0.7
          weight: 0.3
merge_method: ties
base_model: trashpanda-org/MS-24B-Instruct-Mullein-v0
parameters:
  normalize: true
  int8_mask: true
  t:
    - filter: self_attn
      value: [0.4, 0.3, 0.3]  # Matches your weight distribution
    - filter: mlp
      value: [0.4, 0.3, 0.3]  # Same weights for MLP layers
    - value: 0.4  # Default weight for other layers
dtype: bfloat16
tokenizer_source: union
tokenizer_config:
  tokens:
    <|im_start|>:
      source: "h34v7/DansXPantheon-RP-Engine-V1.2-24b-Small-Instruct-Ties-Merge"
    <|im_end|>:
      source: "h34v7/DansXPantheon-RP-Engine-V1.2-24b-Small-Instruct-Ties-Merge"
    <|start_header_id|>:
      source: "trashpanda-org/MS-24B-Instruct-Mullein-v0"
      force: true
    <|end_header_id|>:
      source: "trashpanda-org/MS-24B-Instruct-Mullein-v0"
      force: true
chat_template: auto