TallGemma-Raw / README.md
stereoplegic's picture
Upload folder using huggingface_hub
3f328da verified
---
base_model:
- google/gemma-2b
- google/gemma-1.1-2b-it
- google/codegemma-2b
library_name: transformers
tags:
- mergekit
- merge
---
# tallgemma
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:
* [google/gemma-2b](https://huggingface.co/google/gemma-2b)
* [google/gemma-1.1-2b-it](https://huggingface.co/google/gemma-1.1-2b-it)
* [google/codegemma-2b](https://huggingface.co/google/codegemma-2b)
### Configuration
The following YAML configuration was used to produce this model:
```yaml
slices:
- sources:
- model: google/gemma-2b
layer_range: [0, 1]
- sources:
- model: google/codegemma-2b
layer_range: [0, 1]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [0, 1]
- sources:
- model: google/gemma-2b
layer_range: [1, 2]
- sources:
- model: google/codegemma-2b
layer_range: [1, 2]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [1, 2]
- sources:
- model: google/gemma-2b
layer_range: [2, 3]
- sources:
- model: google/codegemma-2b
layer_range: [2, 3]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [2, 3]
- sources:
- model: google/gemma-2b
layer_range: [3, 4]
- sources:
- model: google/codegemma-2b
layer_range: [3, 4]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [3, 4]
- sources:
- model: google/gemma-2b
layer_range: [4, 5]
- sources:
- model: google/codegemma-2b
layer_range: [4, 5]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [4, 5]
- sources:
- model: google/gemma-2b
layer_range: [5, 6]
- sources:
- model: google/codegemma-2b
layer_range: [5, 6]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [5, 6]
- sources:
- model: google/gemma-2b
layer_range: [6, 7]
- sources:
- model: google/codegemma-2b
layer_range: [6, 7]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [6, 7]
- sources:
- model: google/gemma-2b
layer_range: [7, 8]
- sources:
- model: google/codegemma-2b
layer_range: [7, 8]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [7, 8]
- sources:
- model: google/gemma-2b
layer_range: [8, 9]
- sources:
- model: google/codegemma-2b
layer_range: [8, 9]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [8, 9]
- sources:
- model: google/gemma-2b
layer_range: [9, 10]
- sources:
- model: google/codegemma-2b
layer_range: [9, 10]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [9, 10]
- sources:
- model: google/gemma-2b
layer_range: [10, 11]
- sources:
- model: google/codegemma-2b
layer_range: [10, 11]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [10, 11]
- sources:
- model: google/gemma-2b
layer_range: [11, 12]
- sources:
- model: google/codegemma-2b
layer_range: [11, 12]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [11, 12]
- sources:
- model: google/gemma-2b
layer_range: [12, 13]
- sources:
- model: google/codegemma-2b
layer_range: [12, 13]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [12, 13]
- sources:
- model: google/gemma-2b
layer_range: [13, 14]
- sources:
- model: google/codegemma-2b
layer_range: [13, 14]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [13, 14]
- sources:
- model: google/gemma-2b
layer_range: [14, 15]
- sources:
- model: google/codegemma-2b
layer_range: [14, 15]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [14, 15]
- sources:
- model: google/gemma-2b
layer_range: [15, 16]
- sources:
- model: google/codegemma-2b
layer_range: [15, 16]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [15, 16]
- sources:
- model: google/gemma-2b
layer_range: [16, 17]
- sources:
- model: google/codegemma-2b
layer_range: [16, 17]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [16, 17]
- sources:
- model: google/gemma-2b
layer_range: [17, 18]
- sources:
- model: google/codegemma-2b
layer_range: [17, 18]
- sources:
- model: google/gemma-1.1-2b-it
layer_range: [17, 18]
merge_method: passthrough
dtype: float16
```