|
--- |
|
base_model: |
|
- meta-llama/Llama-3.1-8B-Instruct |
|
library_name: transformers |
|
tags: |
|
- mergekit |
|
- merge |
|
license: llama3.1 |
|
--- |
|
# LLaMa-3.1-Instruct-Interleaved-Zeroed-13B |
|
|
|
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: |
|
* [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) |
|
|
|
### Configuration |
|
|
|
The following YAML configuration was used to produce this model: |
|
|
|
```yaml |
|
merge_method: passthrough |
|
dtype: bfloat16 |
|
slices: |
|
# Original first 10 layers (L0-L9) |
|
- sources: |
|
- layer_range: [0, 10] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
|
|
# Skip Dupe A of L10 |
|
# Original L10 |
|
- sources: |
|
- layer_range: [10, 11] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L10 |
|
- sources: |
|
- layer_range: [10, 11] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L11 |
|
- sources: |
|
- layer_range: [11, 12] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L11 |
|
- sources: |
|
- layer_range: [11, 12] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L11 |
|
- sources: |
|
- layer_range: [11, 12] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L12 |
|
- sources: |
|
- layer_range: [12, 13] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L12 |
|
- sources: |
|
- layer_range: [12, 13] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L12 |
|
- sources: |
|
- layer_range: [12, 13] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L13 |
|
- sources: |
|
- layer_range: [13, 14] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L13 |
|
- sources: |
|
- layer_range: [13, 14] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L13 |
|
- sources: |
|
- layer_range: [13, 14] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L14 |
|
- sources: |
|
- layer_range: [14, 15] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L14 |
|
- sources: |
|
- layer_range: [14, 15] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L14 |
|
- sources: |
|
- layer_range: [14, 15] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L15 |
|
- sources: |
|
- layer_range: [15, 16] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L15 |
|
- sources: |
|
- layer_range: [15, 16] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L15 |
|
- sources: |
|
- layer_range: [15, 16] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L16 |
|
- sources: |
|
- layer_range: [16, 17] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L16 |
|
- sources: |
|
- layer_range: [16, 17] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L16 |
|
- sources: |
|
- layer_range: [16, 17] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L17 |
|
- sources: |
|
- layer_range: [17, 18] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L17 |
|
- sources: |
|
- layer_range: [17, 18] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L17 |
|
- sources: |
|
- layer_range: [17, 18] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L18 |
|
- sources: |
|
- layer_range: [18, 19] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L18 |
|
- sources: |
|
- layer_range: [18, 19] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L18 |
|
- sources: |
|
- layer_range: [18, 19] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L19 |
|
- sources: |
|
- layer_range: [19, 20] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L19 |
|
- sources: |
|
- layer_range: [19, 20] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L19 |
|
- sources: |
|
- layer_range: [19, 20] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L20 |
|
- sources: |
|
- layer_range: [20, 21] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L20 |
|
- sources: |
|
- layer_range: [20, 21] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L20 |
|
- sources: |
|
- layer_range: [20, 21] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Dupe A of L21 |
|
- sources: |
|
- layer_range: [21, 22] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
# Original L21 |
|
- sources: |
|
- layer_range: [21, 22] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
# Dupe B of L21 |
|
- sources: |
|
- layer_range: [21, 22] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
parameters: |
|
scale: |
|
- filter: o_proj |
|
value: 0.0 |
|
- filter: down_proj |
|
value: 0.0 |
|
- value: 1.0 |
|
|
|
# Original last 10 layers (L22-L31) |
|
- sources: |
|
- layer_range: [22, 32] |
|
model: meta-llama/Llama-3.1-8B-Instruct |
|
|
|
|
|
``` |
|
|
