Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import tensorflow as tf
|
2 |
+
import tensorflow_hub as hub
|
3 |
+
from tensorflow_text import SentencepieceTokenizer
|
4 |
+
import gradio as gr
|
5 |
+
import math
|
6 |
+
|
7 |
+
model_url = "https://tfhub.dev/google/universal-sentence-encoder-multilingual-large/3"
|
8 |
+
model = hub.load(model_url)
|
9 |
+
|
10 |
+
def embed_text(text: str) -> dict:
|
11 |
+
embeddings = model(text)
|
12 |
+
return embeddings.numpy().tolist()
|
13 |
+
|
14 |
+
embed_text_inter = gr.Interface(
|
15 |
+
fn = embed_text,
|
16 |
+
inputs = "text",
|
17 |
+
outputs = gr.JSON(),
|
18 |
+
title = "Universal Sentence Encoder 3 Large"
|
19 |
+
)
|
20 |
+
|
21 |
+
def distance(text_1: str, text_2: str) -> float:
|
22 |
+
embeddings_1 = model(text_1)
|
23 |
+
embeddings_2 = model(text_2)
|
24 |
+
dist = math.sqrt(sum((embeddings_1 - embeddings_2)**2))
|
25 |
+
return dist
|
26 |
+
|
27 |
+
distance_inter = gr.Interface(
|
28 |
+
fn = distance,
|
29 |
+
inputs = ["text", "text"],
|
30 |
+
outputs = "number",
|
31 |
+
title = "Universal Sentence Encoder 3 Large"
|
32 |
+
)
|
33 |
+
|
34 |
+
|
35 |
+
iface = gr.TabbedInterface(
|
36 |
+
interface_list=[embed_text_inter, distance_inter],
|
37 |
+
title="Universal Sentence Encoder 3 Large"
|
38 |
+
)
|
39 |
+
|
40 |
+
iface.launch()
|