Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -1,25 +1,24 @@
|
|
|
|
1 |
import logging
|
2 |
-
|
3 |
import datasets
|
4 |
-
import
|
5 |
|
6 |
logging.disable(logging.CRITICAL)
|
7 |
|
8 |
-
|
9 |
-
|
10 |
-
)
|
11 |
|
12 |
-
dataset = datasets.load_dataset("json", data_files=["./dataset.json"], split="train")
|
13 |
dataset.load_faiss_index("embeddings", "index.faiss")
|
14 |
|
15 |
-
def search(query
|
16 |
query_embedding = model.encode(query)
|
17 |
_, retrieved_examples = dataset.get_nearest_examples(
|
18 |
"embeddings",
|
19 |
query_embedding,
|
20 |
k=k,
|
21 |
)
|
22 |
-
|
23 |
results = []
|
24 |
for text, start, end, title, url in zip(
|
25 |
retrieved_examples["text"],
|
@@ -32,23 +31,27 @@ def search(query: str, k: int):
|
|
32 |
end = end
|
33 |
result = {
|
34 |
"title": title,
|
35 |
-
"transcript": f"[{str(start)
|
36 |
-
"link": url
|
37 |
}
|
38 |
results.append(result)
|
39 |
return results
|
40 |
|
41 |
iface = gr.Interface(
|
42 |
-
|
43 |
-
inputs=[
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
|
|
|
|
|
|
|
|
|
|
51 |
)
|
52 |
|
53 |
-
|
54 |
-
iface.launch()
|
|
|
1 |
+
import argparse
|
2 |
import logging
|
3 |
+
|
4 |
import datasets
|
5 |
+
import gradio as gr
|
6 |
|
7 |
logging.disable(logging.CRITICAL)
|
8 |
|
9 |
+
model_name = "dangvantuan/sentence-camembert-large"
|
10 |
+
model = gr.load(model_name)
|
|
|
11 |
|
12 |
+
dataset = datasets.load_dataset("json", data_files=["./data/dataset.json"], split="train")
|
13 |
dataset.load_faiss_index("embeddings", "index.faiss")
|
14 |
|
15 |
+
def search(query, k):
|
16 |
query_embedding = model.encode(query)
|
17 |
_, retrieved_examples = dataset.get_nearest_examples(
|
18 |
"embeddings",
|
19 |
query_embedding,
|
20 |
k=k,
|
21 |
)
|
|
|
22 |
results = []
|
23 |
for text, start, end, title, url in zip(
|
24 |
retrieved_examples["text"],
|
|
|
31 |
end = end
|
32 |
result = {
|
33 |
"title": title,
|
34 |
+
"transcript": f"[{str(start)} ====> {str(end)}] {text}",
|
35 |
+
"link": url,
|
36 |
}
|
37 |
results.append(result)
|
38 |
return results
|
39 |
|
40 |
iface = gr.Interface(
|
41 |
+
search,
|
42 |
+
inputs=[
|
43 |
+
gr.inputs.Textbox(label="Query"),
|
44 |
+
gr.inputs.Number(label="K", default=3, min_value=1, max_value=10),
|
45 |
+
],
|
46 |
+
outputs=[
|
47 |
+
gr.outputs.Textbox(label="Title"),
|
48 |
+
gr.outputs.Textbox(label="Transcript"),
|
49 |
+
gr.outputs.Textbox(label="Link"),
|
50 |
+
],
|
51 |
+
title="Camembert and Faiss-powered Search Engine",
|
52 |
+
description="Search through a dataset using Camembert and Faiss",
|
53 |
+
theme="light",
|
54 |
+
layout="vertical",
|
55 |
)
|
56 |
|
57 |
+
iface.launch()
|
|