MrRobotoAI/8b-unaligned-BASE-v2z

merge

This 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 MrRobotoAI/Thor-v1.4-8b-DARK-FICTION as a base.

Models Merged

The following models were included in the merge:

• MrRobotoAI/8b-unaligned-BASE-v2y
• MrRobotoAI/8b-unaligned-BASE-v2m

Configuration

The following YAML configuration was used to produce this model:

merge_method: ties
models:
  - model: MrRobotoAI/8b-unaligned-BASE-v2m
    parameters:
      weight:
        - filter: v_proj
          value: [1, 1, 1, 0.995, 0.995, 0.995, 0.995, 0.995, 1, 1, 1]
        - filter: o_proj
          value: [1, 1, 1, 0.995, 0.995, 0.995, 0.995, 0.995, 1, 1, 1]
        - filter: up_proj
          value: [1, 1, 1, 0.995, 0.995, 0.995, 0.995, 0.995, 1, 1, 1]
        - filter: gate_proj
          value: [1, 1, 1, 0.995, 0.995, 0.995, 0.995, 0.995, 1, 1, 1]
        - filter: down_proj
          value: [1, 1, 1, 0.995, 0.995, 0.995, 0.995, 0.995, 1, 1, 1]
        - value: 1
  - model: MrRobotoAI/8b-unaligned-BASE-v2y
    parameters:
      weight:
        - filter: v_proj
          value: [0, 0, 0, 0.005, 0.005, 0.005, 0.005, 0.005, 0, 0, 0]
        - filter: o_proj
          value: [0, 0, 0, 0.005, 0.005, 0.005, 0.005, 0.005, 0, 0, 0]
        - filter: up_proj
          value: [0, 0, 0, 0.005, 0.005, 0.005, 0.005, 0.005, 0, 0, 0]
        - filter: gate_proj
          value: [0, 0, 0, 0.005, 0.005, 0.005, 0.005, 0.005, 0, 0, 0]
        - filter: down_proj
          value: [0, 0, 0, 0.005, 0.005, 0.005, 0.005, 0.005, 0, 0, 0]
        - value: 0
base_model: MrRobotoAI/Thor-v1.4-8b-DARK-FICTION
tokenizer_source: base
dtype: bfloat16