|
|
--- |
|
|
base_model: |
|
|
- CultriX/MergeStage1v3 |
|
|
- sometimesanotion/Lamarck-14B-v0.7-rc4 |
|
|
library_name: transformers |
|
|
tags: |
|
|
- mergekit |
|
|
- merge |
|
|
|
|
|
--- |
|
|
# merge |
|
|
|
|
|
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 [SLERP](https://en.wikipedia.org/wiki/Slerp) merge method. |
|
|
|
|
|
### Models Merged |
|
|
|
|
|
The following models were included in the merge: |
|
|
* [CultriX/MergeStage1v3](https://huggingface.co/CultriX/MergeStage1v3) |
|
|
* [sometimesanotion/Lamarck-14B-v0.7-rc4](https://huggingface.co/sometimesanotion/Lamarck-14B-v0.7-rc4) |
|
|
|
|
|
### Configuration |
|
|
|
|
|
The following YAML configuration was used to produce this model: |
|
|
|
|
|
```yaml |
|
|
# Stage 2: Slerp with Lamarck Components [Optimized] |
|
|
name: MergeStage2v3 |
|
|
merge_method: slerp |
|
|
base_model: CultriX/MergeStage1v3 |
|
|
tokenizer_source: base # Verify and update if needed |
|
|
dtype: bfloat16 |
|
|
parameters: |
|
|
normalize: true |
|
|
rescale: false |
|
|
int8_mask: true |
|
|
int8_mask: true |
|
|
t: |
|
|
- value: 0.35 # Adjusted starting value |
|
|
slices: |
|
|
- sources: |
|
|
- model: CultriX/MergeStage1v3 |
|
|
layer_range: [0, 6] # Example - Adjust based on model architecture |
|
|
- model: sometimesanotion/Lamarck-14B-v0.7-rc4 |
|
|
layer_range: [0, 6] # Example - Adjust based on model architecture |
|
|
- sources: |
|
|
- model: CultriX/MergeStage1v3 |
|
|
layer_range: [6, 12] # Example - Adjust based on model architecture |
|
|
- model: sometimesanotion/Lamarck-14B-v0.7-rc4 |
|
|
layer_range: [6, 12] # Example - Adjust based on model architecture |
|
|
- sources: |
|
|
- model: CultriX/MergeStage1v3 |
|
|
layer_range: [12, 18] # Example - Adjust based on model architecture |
|
|
- model: sometimesanotion/Lamarck-14B-v0.7-rc4 |
|
|
layer_range: [12, 18] # Example - Adjust based on model architecture |
|
|
- sources: |
|
|
- model: CultriX/MergeStage1v3 |
|
|
layer_range: [18, 24] # Example - Adjust based on model architecture |
|
|
- model: sometimesanotion/Lamarck-14B-v0.7-rc4 |
|
|
layer_range: [18, 24] # Example - Adjust based on model architecture |
|
|
- sources: |
|
|
- model: CultriX/MergeStage1v3 |
|
|
layer_range: [24, 30] # Example - Adjust based on model architecture |
|
|
- model: sometimesanotion/Lamarck-14B-v0.7-rc4 |
|
|
layer_range: [24, 30] # Example - Adjust based on model architecture |
|
|
- sources: |
|
|
- model: CultriX/MergeStage1v3 |
|
|
layer_range: [30, 36] # Example - Adjust based on model architecture |
|
|
- model: sometimesanotion/Lamarck-14B-v0.7-rc4 |
|
|
layer_range: [30, 36] # Example - Adjust based on model architecture |
|
|
- sources: |
|
|
- model: CultriX/MergeStage1v3 |
|
|
layer_range: [36, 42] # Example - Adjust based on model architecture |
|
|
- model: sometimesanotion/Lamarck-14B-v0.7-rc4 |
|
|
layer_range: [36, 42] # Example - Adjust based on model architecture |
|
|
- sources: |
|
|
- model: CultriX/MergeStage1v3 |
|
|
layer_range: [42, 48] # Example - Adjust based on model architecture |
|
|
- model: sometimesanotion/Lamarck-14B-v0.7-rc4 |
|
|
layer_range: [42, 48] # Example - Adjust based on model architectur |
|
|
``` |
|
|
|
