phantasm
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the DARE TIES merge method using kromcomp/L3.1-Spectrev1-12B 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-Spectrev1-12B
dtype: float32
merge_method: dare_ties
modules:
default:
slices:
- sources:
- layer_range: [0, 50]
model: kromcomp/L3.1-Twilightv1-12B
parameters:
density:
- filter: self_attn
value: 0.6
- filter: mlp
value: 0.4
- value: 0.5
lambda:
- filter: self_attn
value: 1.05
- value: 1.0
weight: [0.3, 0.7]
- layer_range: [0, 50]
model: kromcomp/L3.1-Spectrev1-12B
parameters:
density:
- filter: self_attn
value: 0.4
- filter: mlp
value: 0.6
- value: 0.5
lambda: 1.0
weight: [0.7, 0.3]
parameters:
normalize: 1.0
tokenizer:
pad_to_multiple_of: 32
- Downloads last month
- 119
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
๐
Ask for provider support