Create app.py
Browse files
app.py
CHANGED
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# app.py
|
2 |
+
import streamlit as st
|
3 |
+
from transformers import pipeline, TFAutoModelForSequenceClassification, AutoTokenizer
|
4 |
+
from datasets import load_dataset
|
5 |
+
import pandas as pd
|
6 |
+
|
7 |
+
st.set_page_config(layout="wide")
|
8 |
+
|
9 |
+
# Load dataset
|
10 |
+
@st.cache_resource
|
11 |
+
def load_data():
|
12 |
+
dataset = load_dataset("WhiteAngelss/Turkce-Duygu-Analizi-Dataset")
|
13 |
+
return dataset
|
14 |
+
|
15 |
+
dataset = load_data()
|
16 |
+
st.title("Sentiment Analysis with Turkish Dataset")
|
17 |
+
|
18 |
+
# Show dataset examples
|
19 |
+
st.subheader("Sample Data")
|
20 |
+
sample_df = pd.DataFrame(dataset['train'])
|
21 |
+
st.write(sample_df.head())
|
22 |
+
|
23 |
+
# Set up model
|
24 |
+
model_list = ['WhiteAngelss/entity-word-sentiment-analysis']
|
25 |
+
st.sidebar.header("Select Sentiment Analysis Model")
|
26 |
+
model_checkpoint = st.sidebar.radio("", model_list)
|
27 |
+
|
28 |
+
st.sidebar.write("For details of models: 'https://huggingface.co/WhiteAngelss/entity-word-sentiment-analysis'")
|
29 |
+
st.sidebar.write("")
|
30 |
+
|
31 |
+
st.subheader("Select Text Input Method")
|
32 |
+
input_method = st.radio("", ('Select from Examples', 'Write or Paste New Text'))
|
33 |
+
|
34 |
+
if input_method == 'Select from Examples':
|
35 |
+
example_texts = dataset['train']['text'][:5] # Sample examples from the dataset
|
36 |
+
selected_text = st.selectbox('Select Text from List', example_texts)
|
37 |
+
st.subheader("Text to Analyze")
|
38 |
+
input_text = st.text_area("Selected Text", selected_text, height=128, max_chars=None)
|
39 |
+
elif input_method == "Write or Paste New Text":
|
40 |
+
st.subheader("Text to Analyze")
|
41 |
+
input_text = st.text_area('Write or Paste Text Below', value="", height=128, max_chars=None)
|
42 |
+
|
43 |
+
@st.cache_resource
|
44 |
+
def setModel(model_checkpoint):
|
45 |
+
model = TFAutoModelForSequenceClassification.from_pretrained(model_checkpoint)
|
46 |
+
tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)
|
47 |
+
return pipeline('sentiment-analysis', model=model, tokenizer=tokenizer)
|
48 |
+
|
49 |
+
Run_Button = st.button("Run", key=None)
|
50 |
+
|
51 |
+
if Run_Button and input_text:
|
52 |
+
sentiment_pipeline = setModel(model_checkpoint)
|
53 |
+
output = sentiment_pipeline(input_text)
|
54 |
+
|
55 |
+
st.subheader("Sentiment Analysis Results")
|
56 |
+
df = pd.DataFrame(output)
|
57 |
+
st.dataframe(df)
|
58 |
+
|
59 |
+
# Display the sentiment in a more user-friendly format
|
60 |
+
sentiment = output[0]['label']
|
61 |
+
score = output[0]['score']
|
62 |
+
st.write(f"Sentiment: {sentiment} (Score: {score:.2f})")
|