Llama-3-Steerpike-v1-OAS-8B

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

This model might result in characters who are "too" smart if conversation veers into the analytical, but that may be fine depending on the context. There are issues early on with the consistency of formatting, though that will stabilize with more context. This model is imperfect, but interesting.

Tested lightly with Instruct prompts, minP=0.01, and temperature 1+.

Built with Meta Llama 3.

Merge Details

Merge Method

This model was merged using the task arithmetic merge method using mlabonne/NeuralDaredevil-8B-abliterated as a base.

Models Merged

The following models were included in the merge:

• Hastagaras/Halu-OAS-8B-Llama3
• openlynn/Llama-3-Soliloquy-8B-v2
• grimjim/llama-3-aaditya-OpenBioLLM-8B
• NeverSleep/Llama-3-Lumimaid-8B-v0.1-OAS

Configuration

The following YAML configuration was used to produce this model:

base_model: mlabonne/NeuralDaredevil-8B-abliterated
dtype: bfloat16
merge_method: task_arithmetic
slices:
- sources:
  - layer_range: [0, 32]
    model: mlabonne/NeuralDaredevil-8B-abliterated
  - layer_range: [0, 32]
    model: NeverSleep/Llama-3-Lumimaid-8B-v0.1-OAS
    parameters:
      weight: 0.5
  - layer_range: [0, 32]
    model: Hastagaras/Halu-OAS-8B-Llama3
    parameters:
      weight: 0.2
  - layer_range: [0, 32]
    model: openlynn/Llama-3-Soliloquy-8B-v2
    parameters:
      weight: 0.03
  - layer_range: [0, 32]
    model: grimjim/llama-3-aaditya-OpenBioLLM-8B
    parameters:
      weight: 0.1