Commit
·
2c94baf
1
Parent(s):
7228eb6
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,40 @@
|
|
1 |
---
|
2 |
license: mit
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: mit
|
3 |
+
language:
|
4 |
+
- en
|
5 |
+
tags:
|
6 |
+
- NoSleep
|
7 |
+
- Reddit
|
8 |
+
- Story
|
9 |
+
- Horror
|
10 |
+
widget:
|
11 |
+
- text: "[WP] "
|
12 |
+
example_title: "[WP] "
|
13 |
+
datasets:
|
14 |
+
- chloeliu/reddit_nosleep_posts
|
15 |
---
|
16 |
+
|
17 |
+
# "NoSleep" Writing Prompt Generator
|
18 |
+
Finetuned version of [GPT2](https://huggingface.co/gpt2) to facilitate generation of Writing Prompts for the [GPT-NoSleep-355m model](https://huggingface.co/DarwinAnim8or/GPT-NoSleep-355m)
|
19 |
+
|
20 |
+
You can use the space linked on the right to use this model, then use the NoSleep model in tandem to generate stories!
|
21 |
+
|
22 |
+
# Training Procedure
|
23 |
+
This was trained on the 'reddt-nosleep-posts' dataset, using the "HappyTransformers" library on Google Colab.
|
24 |
+
This model was trained for X epochs with learning rate 1e-2.
|
25 |
+
|
26 |
+
# Biases & Limitations
|
27 |
+
This likely contains the same biases and limitations as the original GPT2 that it is based on, and additionally heavy biases from the dataset.
|
28 |
+
It likely will generate offensive output.
|
29 |
+
|
30 |
+
# Intended Use
|
31 |
+
This model is meant for fun, nothing else.
|
32 |
+
|
33 |
+
# Sample Use
|
34 |
+
```python
|
35 |
+
from happytransformer import GENSettings
|
36 |
+
args_top_k = GENSettings(no_repeat_ngram_size=1, do_sample=True, top_k=80, temperature=0.4, max_length=25, early_stopping=True)
|
37 |
+
|
38 |
+
result = happy_gen.generate_text("[WP] \"", args=args_top_k)
|
39 |
+
print(result.text)
|
40 |
+
```
|