metadata
base_model: []
library_name: transformers
tags:
- mergekit
- merge
Fett-uccine-Long-Noodle-7B-120k-Context
This is a merge of pre-trained language models created using mergekit.
Merge Details
A merge with Fett-uccine and Mistral Yarn 120k ctx.
Credit to Nitral for the merge script and idea.
Merge Method
This model was merged using the SLERP merge method.
Models Merged
The following models were included in the merge:
- Z:\ModelColdStorage\Yarn-Mistral-7b-128k
- Z:\ModelColdStorage\Fett-uccine-7B
Configuration
The following YAML configuration was used to produce this model:
slices:
- sources:
- model: Z:\ModelColdStorage\Fett-uccine-7B
layer_range: [0, 32]
- model: Z:\ModelColdStorage\Yarn-Mistral-7b-128k
layer_range: [0, 32]
merge_method: slerp
base_model: Z:\ModelColdStorage\Fett-uccine-7B
parameters:
t:
- filter: self_attn
value: [0, 0.5, 0.3, 0.7, 1]
- filter: mlp
value: [1, 0.5, 0.7, 0.3, 0]
- value: 0.5
dtype: bfloat16