Spaces:
Running
Running
add colab notebook
Browse files- README.md +1 -1
- app.py +30 -31
- notebooks/gradio_app.ipynb +99 -0
README.md
CHANGED
@@ -10,7 +10,7 @@ pinned: false
|
|
10 |
license: gpl-3.0
|
11 |
---
|
12 |
|
13 |
-
# inBERTolate
|
14 |
## Hit your word count by using BERT to pad out your essays!
|
15 |
|
16 |
Sentences are generated that are in context with both the preceding and following sentences. Models like GPT are not well suited to this task as they are Causal Language Models, or autoregressive models, that generate tokens from left to right, conditional on the text that has come before. The B in BERT, on the other hand, stands for "Bidirectional" and it was trained to be able to fill in the gaps using context on either side. BERT is an example of an autoencoder model.
|
|
|
10 |
license: gpl-3.0
|
11 |
---
|
12 |
|
13 |
+
# inBERTolate [![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/teticio/inBERTolate/blob/master/notebooks/gradio_app.ipynb)
|
14 |
## Hit your word count by using BERT to pad out your essays!
|
15 |
|
16 |
Sentences are generated that are in context with both the preceding and following sentences. Models like GPT are not well suited to this task as they are Causal Language Models, or autoregressive models, that generate tokens from left to right, conditional on the text that has come before. The B in BERT, on the other hand, stands for "Bidirectional" and it was trained to be able to fill in the gaps using context on either side. BERT is an example of an autoencoder model.
|
app.py
CHANGED
@@ -164,41 +164,40 @@ def inbertolate(doc: str,
|
|
164 |
new_doc += '\n'
|
165 |
return new_doc
|
166 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
167 |
|
168 |
if __name__ == '__main__':
|
169 |
parser = argparse.ArgumentParser()
|
170 |
parser.add_argument('--port', type=int)
|
171 |
parser.add_argument('--server', type=int)
|
172 |
args = parser.parse_args()
|
173 |
-
|
174 |
-
demo = gr.Interface(
|
175 |
-
fn=inbertolate,
|
176 |
-
title="inBERTolate",
|
177 |
-
description=f"Hit your word count by using BERT ({pretrained}) to pad out your essays!",
|
178 |
-
inputs=[
|
179 |
-
gr.Textbox(label="Text", lines=10),
|
180 |
-
gr.Slider(label="Maximum length to insert between sentences",
|
181 |
-
minimum=1,
|
182 |
-
maximum=40,
|
183 |
-
step=1,
|
184 |
-
value=max_len),
|
185 |
-
gr.Slider(label="Top k", minimum=0, maximum=200, value=top_k),
|
186 |
-
gr.Slider(label="Temperature",
|
187 |
-
minimum=0,
|
188 |
-
maximum=2,
|
189 |
-
value=temperature),
|
190 |
-
gr.Slider(label="Typical p",
|
191 |
-
minimum=0,
|
192 |
-
maximum=1,
|
193 |
-
value=typical_p),
|
194 |
-
gr.Slider(label="Maximum iterations",
|
195 |
-
minimum=0,
|
196 |
-
maximum=1000,
|
197 |
-
value=max_iter),
|
198 |
-
gr.Slider(label="Burn-in",
|
199 |
-
minimum=0,
|
200 |
-
maximum=500,
|
201 |
-
value=burnin),
|
202 |
-
],
|
203 |
-
outputs=gr.Textbox(label="Expanded text", lines=30))
|
204 |
demo.launch(server_name=args.server or '0.0.0.0', server_port=args.port)
|
|
|
164 |
new_doc += '\n'
|
165 |
return new_doc
|
166 |
|
167 |
+
demo = gr.Interface(
|
168 |
+
fn=inbertolate,
|
169 |
+
title="inBERTolate",
|
170 |
+
description=f"Hit your word count by using BERT ({pretrained}) to pad out your essays!",
|
171 |
+
inputs=[
|
172 |
+
gr.Textbox(label="Text", lines=10),
|
173 |
+
gr.Slider(label="Maximum length to insert between sentences",
|
174 |
+
minimum=1,
|
175 |
+
maximum=40,
|
176 |
+
step=1,
|
177 |
+
value=max_len),
|
178 |
+
gr.Slider(label="Top k", minimum=0, maximum=200, value=top_k),
|
179 |
+
gr.Slider(label="Temperature",
|
180 |
+
minimum=0,
|
181 |
+
maximum=2,
|
182 |
+
value=temperature),
|
183 |
+
gr.Slider(label="Typical p",
|
184 |
+
minimum=0,
|
185 |
+
maximum=1,
|
186 |
+
value=typical_p),
|
187 |
+
gr.Slider(label="Maximum iterations",
|
188 |
+
minimum=0,
|
189 |
+
maximum=1000,
|
190 |
+
value=max_iter),
|
191 |
+
gr.Slider(label="Burn-in",
|
192 |
+
minimum=0,
|
193 |
+
maximum=500,
|
194 |
+
value=burnin),
|
195 |
+
],
|
196 |
+
outputs=gr.Textbox(label="Expanded text", lines=30))
|
197 |
|
198 |
if __name__ == '__main__':
|
199 |
parser = argparse.ArgumentParser()
|
200 |
parser.add_argument('--port', type=int)
|
201 |
parser.add_argument('--server', type=int)
|
202 |
args = parser.parse_args()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
203 |
demo.launch(server_name=args.server or '0.0.0.0', server_port=args.port)
|
notebooks/gradio_app.ipynb
ADDED
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cells": [
|
3 |
+
{
|
4 |
+
"cell_type": "markdown",
|
5 |
+
"id": "bd31ba17",
|
6 |
+
"metadata": {},
|
7 |
+
"source": [
|
8 |
+
"<a href=\"https://colab.research.google.com/github/teticio/inBERTolate/blob/master/notebooks/gradio_app.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
|
9 |
+
]
|
10 |
+
},
|
11 |
+
{
|
12 |
+
"cell_type": "code",
|
13 |
+
"execution_count": 1,
|
14 |
+
"id": "f2fa778e",
|
15 |
+
"metadata": {},
|
16 |
+
"outputs": [],
|
17 |
+
"source": [
|
18 |
+
"try:\n",
|
19 |
+
" # are we running on Google Colab?\n",
|
20 |
+
" import google.colab\n",
|
21 |
+
" !git clone -q https://github.com/teticio/audio-diffusion.git\n",
|
22 |
+
" %cd audio-diffusion\n",
|
23 |
+
" !pip install -q -r requirements.txt\n",
|
24 |
+
"except:\n",
|
25 |
+
" pass"
|
26 |
+
]
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"cell_type": "code",
|
30 |
+
"execution_count": 4,
|
31 |
+
"id": "3fd44423",
|
32 |
+
"metadata": {},
|
33 |
+
"outputs": [],
|
34 |
+
"source": [
|
35 |
+
"import os\n",
|
36 |
+
"import sys\n",
|
37 |
+
"sys.path.insert(0, os.path.dirname(os.path.abspath(\"\")))"
|
38 |
+
]
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"cell_type": "code",
|
42 |
+
"execution_count": null,
|
43 |
+
"id": "bb9e5adc",
|
44 |
+
"metadata": {},
|
45 |
+
"outputs": [],
|
46 |
+
"source": [
|
47 |
+
"import app\n",
|
48 |
+
"app.demo.launch(share=True);"
|
49 |
+
]
|
50 |
+
},
|
51 |
+
{
|
52 |
+
"cell_type": "code",
|
53 |
+
"execution_count": null,
|
54 |
+
"id": "e6aefd43",
|
55 |
+
"metadata": {},
|
56 |
+
"outputs": [],
|
57 |
+
"source": []
|
58 |
+
}
|
59 |
+
],
|
60 |
+
"metadata": {
|
61 |
+
"accelerator": "GPU",
|
62 |
+
"colab": {
|
63 |
+
"provenance": []
|
64 |
+
},
|
65 |
+
"gpuClass": "standard",
|
66 |
+
"kernelspec": {
|
67 |
+
"display_name": "huggingface",
|
68 |
+
"language": "python",
|
69 |
+
"name": "huggingface"
|
70 |
+
},
|
71 |
+
"language_info": {
|
72 |
+
"codemirror_mode": {
|
73 |
+
"name": "ipython",
|
74 |
+
"version": 3
|
75 |
+
},
|
76 |
+
"file_extension": ".py",
|
77 |
+
"mimetype": "text/x-python",
|
78 |
+
"name": "python",
|
79 |
+
"nbconvert_exporter": "python",
|
80 |
+
"pygments_lexer": "ipython3",
|
81 |
+
"version": "3.10.4"
|
82 |
+
},
|
83 |
+
"toc": {
|
84 |
+
"base_numbering": 1,
|
85 |
+
"nav_menu": {},
|
86 |
+
"number_sections": true,
|
87 |
+
"sideBar": true,
|
88 |
+
"skip_h1_title": false,
|
89 |
+
"title_cell": "Table of Contents",
|
90 |
+
"title_sidebar": "Contents",
|
91 |
+
"toc_cell": false,
|
92 |
+
"toc_position": {},
|
93 |
+
"toc_section_display": true,
|
94 |
+
"toc_window_display": false
|
95 |
+
}
|
96 |
+
},
|
97 |
+
"nbformat": 4,
|
98 |
+
"nbformat_minor": 5
|
99 |
+
}
|