Update app.py

app.py

@@ -3,6 +3,33 @@ import streamlit as st
 from textwrap import dedent
 import google.generativeai as genai
 
+from clarifai_grpc.channel.clarifai_channel import ClarifaiChannel
+from clarifai_grpc.grpc.api import resources_pb2, service_pb2, service_pb2_grpc
+from clarifai_grpc.grpc.api.status.status_code_pb2 import SUCCESS
+from PIL import Image
+from io import BytesIO
+from nltk.tokenize import sent_tokenize
+import numpy as np
+
+# Ensure nltk punkt tokenizer data is downloaded
+import nltk
+nltk.download('punkt')
+
+# Constants for image and audio generation
+USER_ID_IMG = 'openai'
+APP_ID_IMG = 'dall-e'
+MODEL_ID_IMG = 'dall-e-3'
+MODEL_VERSION_ID_IMG = 'dc9dcb6ee67543cebc0b9a025861b868'
+
+USER_ID_AUDIO = 'eleven-labs'
+APP_ID_AUDIO = 'audio-generation'
+MODEL_ID_AUDIO = 'speech-synthesis'
+MODEL_VERSION_ID_AUDIO = 'f2cead3a965f4c419a61a4a9b501095c'
+
+# Retrieve PAT from environment variable
+PAT = os.getenv('CLARIFAI_PAT')
+
+
 # Tool import
 from crewai.tools.gemini_tools import GeminiSearchTools
 
@@ -16,7 +43,64 @@ from crewai import Agent, Task, Crew, Process
 # Retrieve API Key from Environment Variable
 GOOGLE_AI_STUDIO = os.environ.get('GOOGLE_API_KEY')
 
-image=""
+# Story book
+
+# Function to generate image using Clarifai
+def generate_image(prompt):
+    channel = ClarifaiChannel.get_grpc_channel()
+    stub = service_pb2_grpc.V2Stub(channel)
+    metadata = (('authorization', 'Key ' + PAT),)
+    userDataObject = resources_pb2.UserAppIDSet(user_id=USER_ID_IMG, app_id=APP_ID_IMG)
+
+    post_model_outputs_response = stub.PostModelOutputs(
+        service_pb2.PostModelOutputsRequest(
+            user_app_id=userDataObject,
+            model_id=MODEL_ID_IMG,
+            version_id=MODEL_VERSION_ID_IMG,
+            inputs=[resources_pb2.Input(data=resources_pb2.Data(text=resources_pb2.Text(raw=prompt)))]
+        ),
+        metadata=metadata
+    )
+
+    if post_model_outputs_response.status.code != SUCCESS:
+        return None, "Error in generating image: " + post_model_outputs_response.status.description
+    else:
+        output = post_model_outputs_response.outputs[0].data.image.base64
+        image = Image.open(BytesIO(output))
+        return image, None
+
+# Function to generate audio using Clarifai
+def generate_audio(prompt):
+    channel = ClarifaiChannel.get_grpc_channel()
+    stub = service_pb2_grpc.V2Stub(channel)
+    metadata = (('authorization', 'Key ' + PAT),)
+    userDataObject = resources_pb2.UserAppIDSet(user_id=USER_ID_AUDIO, app_id=APP_ID_AUDIO)
+
+    response = stub.PostModelOutputs(
+        service_pb2.PostModelOutputsRequest(
+            user_app_id=userDataObject,
+            model_id=MODEL_ID_AUDIO,
+            version_id=MODEL_VERSION_ID_AUDIO,
+            inputs=[resources_pb2.Input(data=resources_pb2.Data(text=resources_pb2.Text(raw=prompt)))]
+        ),
+        metadata=metadata
+    )
+
+    if response.status.code != SUCCESS:
+        return None, "Error in generating audio: " + response.status.description
+    else:
+        audio_output = response.outputs[0].data.audio.base64
+        return audio_output, None
+
+# Function to split text into sentences and then chunk them
+
+def split_text_into_sentences_and_chunks(text, n=8):
+    sentences = sent_tokenize(text)
+    total_sentences = len(sentences)
+    sentences_per_chunk = max(2, total_sentences // n)
+    return [sentences[i:i + sentences_per_chunk] for i in range(0, total_sentences, sentences_per_chunk)]
+
+
 
 # Ensure the API key is available
 if not GOOGLE_AI_STUDIO:
@@ -27,7 +111,7 @@ else:
 
     # Base Example with Gemini Search
 
-    TITLE1 = """<h1 align="center">Clarifai - Crewai Framework</h1>"""
+    TITLE1 = """<h1 align="center">Clarifai NextGen Hackathon</h1>"""
 
 def crewai_process(research_topic):
     # Define your agents with roles and goals
@@ -200,7 +284,7 @@ def crewai_process(research_topic):
 import streamlit as st
 
 # Create tabs
-tab1, tab2, tab3 = st.tabs(["Create Your Story", "Data Visualization", "User Input and Results"])
+tab1, tab2, tab3 = st.tabs(["Create Your Story Script", "Build Your Image/Audio Book", "Interact with Your Charaters"])
 
 # Tab 1: Introduction
 with tab1:
@@ -220,8 +304,34 @@ with tab1:
 
 # Tab 2: Data Visualization
 with tab2:
-    st.header("Data Visualization")
-    st.write("Charts and graphs will be displayed here.")
+    # Streamlit page configuration
+    st.set_page_config(page_title='DALL-E Image and Audio Generator', layout='wide')
+    
+    # Streamlit sidebar elements
+    st.sidebar.title("DALL-E Image and Audio Generator")
+    text_block = st.sidebar.text_area("Enter your text block:")
+    
+    # Streamlit main page
+    st.title('Generate Images and Audio with DALL-E and Clarifai')
+    if st.sidebar.button("Generate Images and Audio"):
+        sentence_chunks = split_text_into_sentences_and_chunks(text_block, 8)
+        prompts = [' '.join(chunk) for chunk in sentence_chunks]
+        cols = st.columns(4)
+        with st.spinner('Generating Content...'):
+            for i, prompt in enumerate(prompts):
+                image, img_error = generate_image(prompt)
+                audio, audio_error = generate_audio(prompt)
+    
+                with cols[i % 4]:
+                    if img_error:
+                        st.error(img_error)
+                    else:
+                        st.image(image, prompt, use_column_width=True)
+    
+                    if audio_error:
+                        st.error(audio_error)
+                    else:
+                        st.audio(audio, format='audio/wav')
 
 # Tab 3: User Input and Results
 with tab3: