|
---
|
|
base_model: []
|
|
library_name: transformers
|
|
tags:
|
|
- mergekit
|
|
- merge
|
|
|
|
---
|
|
# L3-Steno-Maid-Black-LARGE-s1-36-sxy
|
|
|
|
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
|
|
|
## Merge Details
|
|
### Merge Method
|
|
|
|
This model was merged using the passthrough merge method.
|
|
|
|
### Models Merged
|
|
|
|
The following models were included in the merge:
|
|
* G:/7B/L3-Lumimaid-8B-v0.1-OAS
|
|
* G:/7B/L3-Jamet-8B-MK.V-Blackroot
|
|
* G:/7B/L3-8B-Stheno-v3.2
|
|
|
|
### Configuration
|
|
|
|
The following YAML configuration was used to produce this model:
|
|
|
|
```yaml
|
|
# 32 layers -> VS 40
|
|
#models:
|
|
# - model: G:/7B/L3-8B-Stheno-v3.2
|
|
# - model: G:/7B/Llama-3-Lumimaid-8B-v0.1-OAS
|
|
# - model: G:/7B/L3-Jamet-8B-MK.V-Blackroot
|
|
#merge_method: model_stock
|
|
#base_model: G:/7B/L3-8B-Stheno-v3.2
|
|
#dtype: float32
|
|
|
|
slices:
|
|
- sources:
|
|
- model: G:/7B/L3-8B-Stheno-v3.2
|
|
layer_range: [0, 14]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 1
|
|
- filter: down_proj
|
|
value: 1
|
|
- value: 1
|
|
- sources:
|
|
- model: G:/7B/L3-Lumimaid-8B-v0.1-OAS
|
|
layer_range: [8, 20]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 1
|
|
- filter: down_proj
|
|
value: 1
|
|
- value: 1
|
|
- sources:
|
|
- model: G:/7B/L3-Jamet-8B-MK.V-Blackroot
|
|
layer_range: [12, 24]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 1
|
|
- filter: down_proj
|
|
value: 1
|
|
- value: 1
|
|
- sources:
|
|
- model: G:/7B/L3-8B-Stheno-v3.2
|
|
layer_range: [14, 20]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: .8
|
|
- filter: down_proj
|
|
value: .8
|
|
- value: .8
|
|
- sources:
|
|
- model: G:/7B/L3-8B-Stheno-v3.2
|
|
layer_range: [20, 25]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: .8
|
|
- filter: down_proj
|
|
value: .8
|
|
- value: .8
|
|
- sources:
|
|
- model: G:/7B/L3-8B-Stheno-v3.2
|
|
layer_range: [25, 27]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: .6
|
|
- filter: down_proj
|
|
value: .6
|
|
- value: 1
|
|
- sources:
|
|
- model: G:/7B/L3-8B-Stheno-v3.2
|
|
layer_range: [27, 28]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: .9
|
|
- filter: down_proj
|
|
value: .9
|
|
- value: 1
|
|
- sources:
|
|
- model: G:/7B/L3-Lumimaid-8B-v0.1-OAS
|
|
layer_range: [20, 25]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 1
|
|
- filter: down_proj
|
|
value: 1
|
|
- value: 1
|
|
- sources:
|
|
- model: G:/7B/L3-Lumimaid-8B-v0.1-OAS
|
|
layer_range: [25, 27]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: .6
|
|
- filter: down_proj
|
|
value: .6
|
|
- value: 1
|
|
- sources:
|
|
- model: G:/7B/L3-Lumimaid-8B-v0.1-OAS
|
|
layer_range: [27, 31]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 1
|
|
- filter: down_proj
|
|
value: 1
|
|
- value: 1
|
|
- sources:
|
|
- model: G:/7B/L3-Jamet-8B-MK.V-Blackroot
|
|
layer_range: [24, 31]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 1
|
|
- filter: down_proj
|
|
value: 1
|
|
- value: 1
|
|
- sources:
|
|
- model: G:/7B/L3-Jamet-8B-MK.V-Blackroot
|
|
layer_range: [31, 32]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 0.3333333333333
|
|
- filter: down_proj
|
|
value: 0.3333333333333
|
|
- value: 0.3333333333333
|
|
- sources:
|
|
- model: G:/7B/L3-Jamet-8B-MK.V-Blackroot
|
|
layer_range: [31, 32]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 0.4444444444444
|
|
- filter: down_proj
|
|
value: 0.4444444444444
|
|
- value: 0.4444444444444
|
|
- sources:
|
|
- model: G:/7B/L3-Jamet-8B-MK.V-Blackroot
|
|
layer_range: [31, 32]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 0.5555555555555
|
|
- filter: down_proj
|
|
value: 0.5555555555555
|
|
- value: 0.5555555555555
|
|
- sources:
|
|
- model: G:/7B/L3-Jamet-8B-MK.V-Blackroot
|
|
layer_range: [31, 32]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 0.6666666666666
|
|
- filter: down_proj
|
|
value: 0.6666666666666
|
|
- value: 0.6666666666666
|
|
- sources:
|
|
- model: G:/7B/L3-Jamet-8B-MK.V-Blackroot
|
|
layer_range: [31, 32]
|
|
parameters:
|
|
scale:
|
|
- filter: o_proj
|
|
value: 0.7777777777777
|
|
- filter: down_proj
|
|
value: 0.7777777777777
|
|
- value: 0.8888888888888
|
|
merge_method: passthrough
|
|
dtype: float16
|
|
```
|
|
|
