MrRobotoAI/MrRoboto-ProLong-8b-v2i

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 MrRobotoAI/MrRoboto-ProLong-8b-v2k as a base.

Models Merged

The following models were included in the merge:

• MrRobotoAI/Thor-v1.4-8b-DARK-FICTION
• MrRobotoAI/MrRoboto-ProLong-8b-v2h

Configuration

The following YAML configuration was used to produce this model:

merge_method: dare_linear
models:
  - model: MrRobotoAI/Thor-v1.4-8b-DARK-FICTION
    parameters:
      weight:
        - filter: v_proj
          value: [0.25, 0.25, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.25, 0.25]
        - filter: o_proj
          value: [0.25, 0.25, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.25, 0.25]
        - filter: up_proj
          value: [0.25, 0.25, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.25, 0.25]
        - filter: gate_proj
          value: [0.25, 0.25, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.25, 0.25]
        - filter: down_proj
          value: [0.25, 0.25, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.25, 0.25]
        - value: 1
  - model: MrRobotoAI/MrRoboto-ProLong-8b-v2h
    parameters:
      weight:
        - filter: v_proj
          value: [0.75, 0.75, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.75, 0.75]
        - filter: o_proj
          value: [0.75, 0.75, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.75, 0.75]
        - filter: up_proj
          value: [0.75, 0.75, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.75, 0.75]
        - filter: gate_proj
          value: [0.75, 0.75, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.75, 0.75]
        - filter: down_proj
          value: [0.75, 0.75, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.75, 0.75]
        - value: 0
base_model: MrRobotoAI/MrRoboto-ProLong-8b-v2k
tokenizer_source: base
dtype: bfloat16