metadata
base_model:
- mistralai/Mistral-7B-Instruct-v0.2
library_name: transformers
tags:
- mergekit
- merge
merge
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the linear merge method.
Models Merged
The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
dtype: float16
merge_method: linear
slices:
- sources:
- layer_range: [0, 8] # Reduce the first half of the model to 4B parameters
model: mistralai/Mistral-7B-Instruct-v0.2
parameters:
weight: 0.25 # Reduce the weight of the first half to make room for the second half
- layer_range: [8, 16] # Reduce the second half of the model to 4B parameters
model: mistralai/Mistral-7B-Instruct-v0.2
parameters:
weight: 0.25 # Maintain the weight of the second half
- layer_range: [16, 24] # Reduce the third half of the model to 4B parameters
model: mistralai/Mistral-7B-Instruct-v0.2
parameters:
weight: 0.25 # Reduce the weight of the third half
- layer_range: [24, 32] # Reduce the fourth half of the model to 4B parameters
model: mistralai/Mistral-7B-Instruct-v0.2
parameters:
weight: 0.25 # Maintain the weight of the fourth half