Miqu Models ('task_arithmetic')
Collection
A collection of creative writing models based on the 'miqu-1-70b ' model.
•
5 items
•
Updated
This is a merge of pre-trained language models created using mergekit.
This model was merged using the linear merge method using /home/juk/LLMs/models/huggingface/miqu-1-70b-sf as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
base_model:
model:
path: /home/juk/LLMs/models/huggingface/miqu-1-70b-sf
dtype: float16
merge_method: linear
slices:
- sources:
- layer_range: [0, 80]
model:
model:
path: /home/juk/LLMs/models/huggingface/miqu-1-70b-sf
parameters:
weight:
- filter: v_proj
value: [1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0]
- filter: o_proj
value: [1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0]
- filter: up_proj
value: [1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0]
- filter: gate_proj
value: [1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0]
- filter: down_proj
value: [1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0]
- value: 1.0
- layer_range: [0, 80]
model:
model:
path: miqu-models/_miquplus-xwin-70b
parameters:
weight:
- filter: v_proj
value: [0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0]
- filter: o_proj
value: [0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0]
- filter: up_proj
value: [0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0]
- filter: gate_proj
value: [0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0]
- filter: down_proj
value: [0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0]
- value: 0.0
tokenizer_source: base