Quantization made by Richard Erkhov.
This is a merge of pre-trained language models created using
mergekit.
Each merge component was already subjected to Orthogonal Activation Steering (OAS) to mitigate refusals. The resulting text completion model should be versatile for both positive and negative roleplay scenarios and storytelling. Care should be taken when using this model.
Built with Meta Llama 3.
This model was merged using the
task arithmetic merge method using
mlabonne/NeuralDaredevil-8B-abliterated as a base.
1base_model: mlabonne/NeuralDaredevil-8B-abliterated
2dtype: bfloat16
3merge_method: task_arithmetic
4slices:
5- sources:
6 - layer_range: [0, 32]
7 model: mlabonne/NeuralDaredevil-8B-abliterated
8 - layer_range: [0, 32]
9 model: NeverSleep/Llama-3-Lumimaid-8B-v0.1-OAS
10 parameters:
11 weight: 0.3
12 - layer_range: [0, 32]
13 model: Hastagaras/Halu-OAS-8B-Llama3
14 parameters:
15 weight: 0.3
16
