Editing Models with Task Arithmetic
Paper
•
2212.04089
•
Published
•
7
This is a merge of pre-trained language models created using mergekit.
This model was merged using the task arithmetic merge method using Delta-Vector/Baldur-8B + NewEden/baldur-DPO-lora as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
base_model: Delta-Vector/Baldur-8B+NewEden/baldur-DPO-lora
dtype: bfloat16
merge_method: task_arithmetic
parameters:
normalize: false
slices:
- sources:
- layer_range: [0, 32]
model: Delta-Vector/Baldur-8B+NewEden/baldur-DPO-lora
parameters:
weight: 1.0
Base model
meta-llama/Llama-3.1-8B