Update app.py
Browse files
app.py
CHANGED
@@ -1,48 +1,103 @@
|
|
|
|
1 |
import os
|
2 |
-
import requests
|
3 |
import streamlit as st
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
4 |
|
5 |
-
#
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
# Input text for sentiment analysis
|
14 |
-
input_text = st.text_area("Enter movie review:", "")
|
15 |
-
|
16 |
-
# Choose analysis type
|
17 |
-
analysis_type = st.radio("Select analysis type:", ["Zero-shot", "One-shot", "Few-shot"])
|
18 |
-
|
19 |
-
if st.button("Analyze Sentiment"):
|
20 |
-
# Prepare payload for API request
|
21 |
-
if analysis_type == "Zero-shot":
|
22 |
-
payload = {"inputs": f"Label the text as either 'positive', 'negative', or 'mixed' related to a movie:\n\n{input_text}"}
|
23 |
-
elif analysis_type == "One-shot":
|
24 |
-
prompt = "Label the sentence as either 'positive', 'negative', or 'mixed' related to a movie:\n\n" \
|
25 |
-
"Sentence: This movie exceeded my expectations.\nLabel: positive"
|
26 |
-
payload = {"inputs": f"{prompt} {input_text}"}
|
27 |
-
elif analysis_type == "Few-shot":
|
28 |
-
examples = [
|
29 |
-
"Sentence: The cinematography in this movie is outstanding.\nLabel: positive",
|
30 |
-
"Sentence: I didn't enjoy the plot twists in the movie.\nLabel: negative",
|
31 |
-
"Sentence: The acting was great, but the pacing felt off.\nLabel: mixed",
|
32 |
-
"Sentence: This movie didn't live up to the hype.\nLabel: negative",
|
33 |
-
]
|
34 |
-
prompt = "Label the sentences as either 'positive', 'negative', or 'mixed' related to a movie:\n\n" + "\n".join(examples)
|
35 |
-
payload = {"inputs": f"{prompt}\n\n{input_text}"}
|
36 |
-
|
37 |
-
# Make API request
|
38 |
-
response = requests.post(API_URL, headers=HEADERS, json=payload)
|
39 |
-
|
40 |
-
# Print entire response for debugging
|
41 |
-
st.write("API Response:", response.json())
|
42 |
-
|
43 |
-
# Display results
|
44 |
-
if response.status_code == 200:
|
45 |
-
result = response.json()[0] # Assuming the sentiment is in the first item of the list
|
46 |
-
st.write("Sentiment:", result.get('generated_text', 'N/A'))
|
47 |
-
else:
|
48 |
-
st.write("Error:", response.status_code, response.text)
|
|
|
1 |
+
# Imports
|
2 |
import os
|
|
|
3 |
import streamlit as st
|
4 |
+
import requests
|
5 |
+
from transformers import pipeline
|
6 |
+
import openai
|
7 |
+
|
8 |
+
# Suppressing all warnings
|
9 |
+
import warnings
|
10 |
+
warnings.filterwarnings("ignore")
|
11 |
+
|
12 |
+
# Image-to-text
|
13 |
+
def img2txt(url):
|
14 |
+
print("Initializing captioning model...")
|
15 |
+
captioning_model = pipeline("image-to-text", model="Salesforce/blip-image-captioning-base")
|
16 |
+
|
17 |
+
print("Generating text from the image...")
|
18 |
+
text = captioning_model(url, max_new_tokens=20)[0]["generated_text"]
|
19 |
+
|
20 |
+
print(text)
|
21 |
+
return text
|
22 |
+
|
23 |
+
# Text-to-story
|
24 |
+
def txt2story(img_text):
|
25 |
+
print("Initializing client...")
|
26 |
+
client = openai.OpenAI(
|
27 |
+
api_key=os.environ["TOGETHER_API_KEY"],
|
28 |
+
base_url='https://api.together.xyz',
|
29 |
+
)
|
30 |
+
|
31 |
+
messages = [
|
32 |
+
{"role": "system", "content": '''As an experienced short story writer, write story title and then create a meaningful story influenced by provided words.
|
33 |
+
Ensure stories conclude positively within 100 words. Remember the story must end within 100 words''', "temperature": 1.8},
|
34 |
+
{"role": "user", "content": f"Here is input set of words: {img_text}", "temperature": 1.5},
|
35 |
+
]
|
36 |
+
|
37 |
+
print("Story...")
|
38 |
+
chat_completion = client.chat.completions.create(
|
39 |
+
messages=messages,
|
40 |
+
model="togethercomputer/llama-2-70b-chat")
|
41 |
+
|
42 |
+
print(chat_completion.choices[0].message.content)
|
43 |
+
return chat_completion.choices[0].message.content
|
44 |
+
|
45 |
+
|
46 |
+
|
47 |
+
# Text-to-speech
|
48 |
+
def txt2speech(text):
|
49 |
+
print("Initializing text-to-speech conversion...")
|
50 |
+
API_URL = "https://api-inference.huggingface.co/models/espnet/kan-bayashi_ljspeech_vits"
|
51 |
+
headers = {"Authorization": f"Bearer {os.environ['HUGGINGFACEHUB_API_TOKEN']}"}
|
52 |
+
payloads = {'inputs': text}
|
53 |
+
|
54 |
+
response = requests.post(API_URL, headers=headers, json=payloads)
|
55 |
+
|
56 |
+
with open('audio_story.mp3', 'wb') as file:
|
57 |
+
file.write(response.content)
|
58 |
+
|
59 |
+
|
60 |
+
# Streamlit web app main function
|
61 |
+
def main():
|
62 |
+
st.set_page_config(page_title="π¨ Image-to-Audio Story π§", page_icon="πΌοΈ")
|
63 |
+
st.title("Turn the Image into Audio Story")
|
64 |
+
|
65 |
+
# Allows users to upload an image file
|
66 |
+
uploaded_file = st.file_uploader("# π· Upload an image...", type=["jpg", "jpeg", "png"])
|
67 |
+
|
68 |
+
if uploaded_file is not None:
|
69 |
+
# Reads and saves uploaded image file
|
70 |
+
bytes_data = uploaded_file.read()
|
71 |
+
with open("uploaded_image.jpg", "wb") as file:
|
72 |
+
file.write(bytes_data)
|
73 |
+
|
74 |
+
st.image(uploaded_file, caption='πΌοΈ Uploaded Image', use_column_width=True)
|
75 |
+
|
76 |
+
# Initiates AI processing and story generation
|
77 |
+
with st.spinner("## π€ AI is at Work! "):
|
78 |
+
scenario = img2txt("uploaded_image.jpg") # Extracts text from the image
|
79 |
+
story = txt2story(scenario) # Generates a story based on the image text
|
80 |
+
txt2speech(story) # Converts the story to audio
|
81 |
+
|
82 |
+
st.markdown("---")
|
83 |
+
st.markdown("## π Image Caption")
|
84 |
+
st.write(scenario)
|
85 |
+
|
86 |
+
st.markdown("---")
|
87 |
+
st.markdown("## π Story")
|
88 |
+
st.write(story)
|
89 |
+
|
90 |
+
st.markdown("---")
|
91 |
+
st.markdown("## π§ Audio Story")
|
92 |
+
st.audio("audio_story.mp3")
|
93 |
+
|
94 |
+
if __name__ == '__main__':
|
95 |
+
main()
|
96 |
|
97 |
+
# Credits
|
98 |
+
st.markdown("### Credits")
|
99 |
+
st.caption('''
|
100 |
+
Made with β€οΈ by @Aditya-Neural-Net-Ninja\n
|
101 |
+
Utilizes Image-to-text, Text Generation, Text-to-speech Transformer Models\n
|
102 |
+
Gratitude to Streamlit, π€ Spaces for Deployment & Hosting
|
103 |
+
''')
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|