mirrorglaze
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the DELLA merge method using kromcomp/L3.1-Nuancev3-12B as a base.
Models Merged
The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
base_model: kromcomp/L3.1-Nuancev3-12B
chat_template: llama3
dtype: float32
merge_method: della
modules:
default:
slices:
- sources:
- layer_range: [0, 50]
model: kromcomp/L3.1-Orotundv1-12B
parameters:
density: 0.6
epsilon: 0.15
lambda: 1.0
weight: [0.55, 0.65, 0.7]
- layer_range: [0, 50]
model: kromcomp/L3.1-Brasseriev1-12B
parameters:
density: 0.3
epsilon: 0.1
lambda: 1.0
weight: [0.35, 0.2, 0.1]
- layer_range: [0, 50]
model: kromcomp/L3.1-Nuancev3-12B
parameters:
density: 0.5
epsilon: 0.1
lambda: 1.0
weight: [0.1, 0.15, 0.2]
parameters:
int8_mask: 1.0
normalize: 0.0
tokenizer:
pad_to_multiple_of: 32
- Downloads last month
- 10
Model tree for kromcomp/L3.1-Mirrorglazev3-12B
Merge model
this model