Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,29 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: mit
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
base_model: openai-community/gpt2
|
| 4 |
+
---
|
| 5 |
+
---
|
| 6 |
+
license: mit
|
| 7 |
+
base_model: openai-community/gpt2-large
|
| 8 |
+
---
|
| 9 |
+
# Null-GPT2
|
| 10 |
+
|
| 11 |
+
## Description
|
| 12 |
+
|
| 13 |
+
This is a GPT2 Model, but only with the architecture, no pre-trained weights, biases, attention, etc.
|
| 14 |
+
|
| 15 |
+
This is useful for researchers who want to play with training the model (not tuning).
|
| 16 |
+
|
| 17 |
+
Generated via the github repo [Model Architecture Generator](https://github.com/ivanhe123/Model-Architecture-Generator)
|
| 18 |
+
|
| 19 |
+
## Use
|
| 20 |
+
First go into the directory of the model,
|
| 21 |
+
```
|
| 22 |
+
git clone https://github.com/ivanhe123/Model-Architecture-Generator
|
| 23 |
+
```
|
| 24 |
+
|
| 25 |
+
```
|
| 26 |
+
python -m randomnize_params -in "./NullGPT2" -out path_model_out
|
| 27 |
+
```
|
| 28 |
+
|
| 29 |
+
path_model_out is just the output path of the newly randomnized model.
|