| | --- |
| | base_model: |
| | - Sao10K/Fimbulvetr-11B-v2 |
| | - TheDrummer/Moistral-11B-v3 |
| | 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 merge method. |
| |
|
| | ### Models Merged |
| |
|
| | The following models were included in the merge: |
| | * [Sao10K/Fimbulvetr-11B-v2](https://huggingface.co/Sao10K/Fimbulvetr-11B-v2) |
| | * [TheDrummer/Moistral-11B-v3](https://huggingface.co/TheDrummer/Moistral-11B-v3) |
| |
|
| | ### Configuration |
| |
|
| | The following YAML configuration was used to produce this model: |
| |
|
| | ```yaml |
| | models: |
| | - model: TheDrummer/Moistral-11B-v3 |
| | - model: Sao10K/Fimbulvetr-11B-v2 |
| | merge_method: slerp |
| | tokenizer_merge_method: slerp |
| | tokenizer_parameters: |
| | t: 0.3 # Dà più peso al tokenizer di Fimbulvetr |
| | base_model: Sao10K/Fimbulvetr-11B-v2 |
| | dtype: bfloat16 |
| | parameters: |
| | t: [0, 0.2, 0.4, 0.5, 0.4, 0.2, 0] # Curva che favorisce leggermente Fimbulvetr |
| | temp: 1.5 # Temperatura per smoothare il merge |
| | density: # Density merging per bilanciare le caratteristiche dei due modelli |
| | - threshold: 0.1 |
| | t: 0.7 |
| | - threshold: 0.5 |
| | t: 0.5 |
| | - threshold: 0.9 |
| | t: 0.3 |
| | ``` |
| |
|
