miquplus-xwin-70b / mergekit_config.yml
jukofyork's picture
Upload mergekit_config.yml with huggingface_hub
7d03496 verified
raw
history blame
1.41 kB
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