Upload folder using huggingface_hub
Browse files- README.md +15 -5
- config.json +1 -1
- mergekit_config.yml +12 -1
- model-00001-of-00001.safetensors +1 -1
README.md
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
---
|
| 2 |
base_model:
|
| 3 |
- mllm-dev/gpt2_f_experiment_4
|
|
|
|
| 4 |
- mllm-dev/gpt2_f_experiment_1
|
| 5 |
- mllm-dev/gpt2_f_experiment_0
|
| 6 |
- mllm-dev/gpt2_f_experiment_3
|
| 7 |
-
- mllm-dev/gpt2_f_experiment_2
|
| 8 |
library_name: transformers
|
| 9 |
tags:
|
| 10 |
- mergekit
|
|
@@ -18,24 +18,29 @@ This is a merge of pre-trained language models created using [mergekit](https://
|
|
| 18 |
## Merge Details
|
| 19 |
### Merge Method
|
| 20 |
|
| 21 |
-
This model was merged using the [
|
| 22 |
|
| 23 |
### Models Merged
|
| 24 |
|
| 25 |
The following models were included in the merge:
|
| 26 |
* [mllm-dev/gpt2_f_experiment_4](https://huggingface.co/mllm-dev/gpt2_f_experiment_4)
|
|
|
|
| 27 |
* [mllm-dev/gpt2_f_experiment_1](https://huggingface.co/mllm-dev/gpt2_f_experiment_1)
|
| 28 |
-
* [mllm-dev/gpt2_f_experiment_0](https://huggingface.co/mllm-dev/gpt2_f_experiment_0)
|
| 29 |
* [mllm-dev/gpt2_f_experiment_3](https://huggingface.co/mllm-dev/gpt2_f_experiment_3)
|
| 30 |
-
* [mllm-dev/gpt2_f_experiment_2](https://huggingface.co/mllm-dev/gpt2_f_experiment_2)
|
| 31 |
|
| 32 |
### Configuration
|
| 33 |
|
| 34 |
The following YAML configuration was used to produce this model:
|
| 35 |
|
| 36 |
```yaml
|
|
|
|
|
|
|
|
|
|
| 37 |
dtype: float16
|
| 38 |
-
merge_method:
|
|
|
|
|
|
|
|
|
|
| 39 |
slices:
|
| 40 |
- sources:
|
| 41 |
- layer_range: [0, 12]
|
|
@@ -43,29 +48,34 @@ slices:
|
|
| 43 |
model:
|
| 44 |
path: mllm-dev/gpt2_f_experiment_0
|
| 45 |
parameters:
|
|
|
|
| 46 |
weight: 1.0
|
| 47 |
- layer_range: [0, 12]
|
| 48 |
model:
|
| 49 |
model:
|
| 50 |
path: mllm-dev/gpt2_f_experiment_1
|
| 51 |
parameters:
|
|
|
|
| 52 |
weight: 1.0
|
| 53 |
- layer_range: [0, 12]
|
| 54 |
model:
|
| 55 |
model:
|
| 56 |
path: mllm-dev/gpt2_f_experiment_2
|
| 57 |
parameters:
|
|
|
|
| 58 |
weight: 1.0
|
| 59 |
- layer_range: [0, 12]
|
| 60 |
model:
|
| 61 |
model:
|
| 62 |
path: mllm-dev/gpt2_f_experiment_3
|
| 63 |
parameters:
|
|
|
|
| 64 |
weight: 1.0
|
| 65 |
- layer_range: [0, 12]
|
| 66 |
model:
|
| 67 |
model:
|
| 68 |
path: mllm-dev/gpt2_f_experiment_4
|
| 69 |
parameters:
|
|
|
|
| 70 |
weight: 1.0
|
| 71 |
```
|
|
|
|
| 1 |
---
|
| 2 |
base_model:
|
| 3 |
- mllm-dev/gpt2_f_experiment_4
|
| 4 |
+
- mllm-dev/gpt2_f_experiment_2
|
| 5 |
- mllm-dev/gpt2_f_experiment_1
|
| 6 |
- mllm-dev/gpt2_f_experiment_0
|
| 7 |
- mllm-dev/gpt2_f_experiment_3
|
|
|
|
| 8 |
library_name: transformers
|
| 9 |
tags:
|
| 10 |
- mergekit
|
|
|
|
| 18 |
## Merge Details
|
| 19 |
### Merge Method
|
| 20 |
|
| 21 |
+
This model was merged using the [DARE](https://arxiv.org/abs/2311.03099) [TIES](https://arxiv.org/abs/2306.01708) merge method using [mllm-dev/gpt2_f_experiment_0](https://huggingface.co/mllm-dev/gpt2_f_experiment_0) as a base.
|
| 22 |
|
| 23 |
### Models Merged
|
| 24 |
|
| 25 |
The following models were included in the merge:
|
| 26 |
* [mllm-dev/gpt2_f_experiment_4](https://huggingface.co/mllm-dev/gpt2_f_experiment_4)
|
| 27 |
+
* [mllm-dev/gpt2_f_experiment_2](https://huggingface.co/mllm-dev/gpt2_f_experiment_2)
|
| 28 |
* [mllm-dev/gpt2_f_experiment_1](https://huggingface.co/mllm-dev/gpt2_f_experiment_1)
|
|
|
|
| 29 |
* [mllm-dev/gpt2_f_experiment_3](https://huggingface.co/mllm-dev/gpt2_f_experiment_3)
|
|
|
|
| 30 |
|
| 31 |
### Configuration
|
| 32 |
|
| 33 |
The following YAML configuration was used to produce this model:
|
| 34 |
|
| 35 |
```yaml
|
| 36 |
+
base_model:
|
| 37 |
+
model:
|
| 38 |
+
path: mllm-dev/gpt2_f_experiment_0
|
| 39 |
dtype: float16
|
| 40 |
+
merge_method: dare_ties
|
| 41 |
+
parameters:
|
| 42 |
+
int8_mask: 1.0
|
| 43 |
+
normalize: 1.0
|
| 44 |
slices:
|
| 45 |
- sources:
|
| 46 |
- layer_range: [0, 12]
|
|
|
|
| 48 |
model:
|
| 49 |
path: mllm-dev/gpt2_f_experiment_0
|
| 50 |
parameters:
|
| 51 |
+
density: 1.0
|
| 52 |
weight: 1.0
|
| 53 |
- layer_range: [0, 12]
|
| 54 |
model:
|
| 55 |
model:
|
| 56 |
path: mllm-dev/gpt2_f_experiment_1
|
| 57 |
parameters:
|
| 58 |
+
density: 1.0
|
| 59 |
weight: 1.0
|
| 60 |
- layer_range: [0, 12]
|
| 61 |
model:
|
| 62 |
model:
|
| 63 |
path: mllm-dev/gpt2_f_experiment_2
|
| 64 |
parameters:
|
| 65 |
+
density: 1.0
|
| 66 |
weight: 1.0
|
| 67 |
- layer_range: [0, 12]
|
| 68 |
model:
|
| 69 |
model:
|
| 70 |
path: mllm-dev/gpt2_f_experiment_3
|
| 71 |
parameters:
|
| 72 |
+
density: 1.0
|
| 73 |
weight: 1.0
|
| 74 |
- layer_range: [0, 12]
|
| 75 |
model:
|
| 76 |
model:
|
| 77 |
path: mllm-dev/gpt2_f_experiment_4
|
| 78 |
parameters:
|
| 79 |
+
density: 1.0
|
| 80 |
weight: 1.0
|
| 81 |
```
|
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "mllm-dev/
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"architectures": [
|
| 5 |
"GPT2ForSequenceClassification"
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "mllm-dev/gpt2_f_experiment_0",
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"architectures": [
|
| 5 |
"GPT2ForSequenceClassification"
|
mergekit_config.yml
CHANGED
|
@@ -1,5 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
| 1 |
dtype: float16
|
| 2 |
-
merge_method:
|
|
|
|
|
|
|
|
|
|
| 3 |
slices:
|
| 4 |
- sources:
|
| 5 |
- layer_range: [0, 12]
|
|
@@ -7,28 +13,33 @@ slices:
|
|
| 7 |
model:
|
| 8 |
path: mllm-dev/gpt2_f_experiment_0
|
| 9 |
parameters:
|
|
|
|
| 10 |
weight: 1.0
|
| 11 |
- layer_range: [0, 12]
|
| 12 |
model:
|
| 13 |
model:
|
| 14 |
path: mllm-dev/gpt2_f_experiment_1
|
| 15 |
parameters:
|
|
|
|
| 16 |
weight: 1.0
|
| 17 |
- layer_range: [0, 12]
|
| 18 |
model:
|
| 19 |
model:
|
| 20 |
path: mllm-dev/gpt2_f_experiment_2
|
| 21 |
parameters:
|
|
|
|
| 22 |
weight: 1.0
|
| 23 |
- layer_range: [0, 12]
|
| 24 |
model:
|
| 25 |
model:
|
| 26 |
path: mllm-dev/gpt2_f_experiment_3
|
| 27 |
parameters:
|
|
|
|
| 28 |
weight: 1.0
|
| 29 |
- layer_range: [0, 12]
|
| 30 |
model:
|
| 31 |
model:
|
| 32 |
path: mllm-dev/gpt2_f_experiment_4
|
| 33 |
parameters:
|
|
|
|
| 34 |
weight: 1.0
|
|
|
|
| 1 |
+
base_model:
|
| 2 |
+
model:
|
| 3 |
+
path: mllm-dev/gpt2_f_experiment_0
|
| 4 |
dtype: float16
|
| 5 |
+
merge_method: dare_ties
|
| 6 |
+
parameters:
|
| 7 |
+
int8_mask: 1.0
|
| 8 |
+
normalize: 1.0
|
| 9 |
slices:
|
| 10 |
- sources:
|
| 11 |
- layer_range: [0, 12]
|
|
|
|
| 13 |
model:
|
| 14 |
path: mllm-dev/gpt2_f_experiment_0
|
| 15 |
parameters:
|
| 16 |
+
density: 1.0
|
| 17 |
weight: 1.0
|
| 18 |
- layer_range: [0, 12]
|
| 19 |
model:
|
| 20 |
model:
|
| 21 |
path: mllm-dev/gpt2_f_experiment_1
|
| 22 |
parameters:
|
| 23 |
+
density: 1.0
|
| 24 |
weight: 1.0
|
| 25 |
- layer_range: [0, 12]
|
| 26 |
model:
|
| 27 |
model:
|
| 28 |
path: mllm-dev/gpt2_f_experiment_2
|
| 29 |
parameters:
|
| 30 |
+
density: 1.0
|
| 31 |
weight: 1.0
|
| 32 |
- layer_range: [0, 12]
|
| 33 |
model:
|
| 34 |
model:
|
| 35 |
path: mllm-dev/gpt2_f_experiment_3
|
| 36 |
parameters:
|
| 37 |
+
density: 1.0
|
| 38 |
weight: 1.0
|
| 39 |
- layer_range: [0, 12]
|
| 40 |
model:
|
| 41 |
model:
|
| 42 |
path: mllm-dev/gpt2_f_experiment_4
|
| 43 |
parameters:
|
| 44 |
+
density: 1.0
|
| 45 |
weight: 1.0
|
model-00001-of-00001.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 248902264
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:11322e145e7b61665593903f460c972df3374b662f8ac11f087211938f7fd91c
|
| 3 |
size 248902264
|