Model Stock: All we need is just a few fine-tuned models
Paper
β’
2403.19522
β’
Published
β’
13
This is a merge of pre-trained language models created using mergekit.
This is a multi-stage merge. There's little method to my madness and I just stopped when I arrived at something that I liked.
Starting point was DepravedCartographer-v1.0-24b with slight changes.
v1.1
v1.2
v1.3
v1.4
This model was merged using the Model Stock merge method using TheDrummer/Cydonia-24B-v4 as a base.
This model was merged using the SLERP merge method.
This model was merged using the NuSLERP merge method using /intermediate/model/B as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
base_model: TheDrummer/Cydonia-24B-v4 # Cydonia v4
merge_method: model_stock
dtype: bfloat16
models:
- model: aixonlab/Eurydice-24b-v3.5 # storytelling / RP
- model: TheDrummer/Cydonia-24B-v4 # sprinkle in some extra Cydonia v4
- model: PocketDoc/Dans-PersonalityEngine-V1.3.0-24b # Prompt Adherence
- model: Delta-Vector/MS3.2-Austral-Winton # Adventure
- model: Doctor-Shotgun/MS3.1-24B-Magnum-Diamond # claude opus
β /intermediate/model/A β
merge_method: slerp
dtype: bfloat16
base_model: anthracite-core/Mistral-Small-3.2-24B-Instruct-2506-Text-Only
models:
- model: /intermediate/model/A
parameters:
t: 0.4
β /intermediate/model/B β
merge_method: nuslerp
dtype: bfloat16
base_model: /intermediate/model/B
models:
- model: Doctor-Shotgun/MS3.2-24B-Magnum-Diamond
parameters:
weight: 0.6
- model: Delta-Vector/MS3.2-Austral-Winton
parameters:
weight: 0.4
β /intermediate/model/C β
merge_method: slerp
dtype: bfloat16
base_model: /intermediate/model/B
models:
- model: /intermediate/model/C
parameters:
t: 0.5
β WeirdCompound-v1.4-24b