import streamlit as st
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 import status_code_pb2
# Set your Clarifai credentials and model details for DALL-E
PAT_DALLE = 'bfdeb4029ef54d23a2e608b0aa4c00e4'
USER_ID_DALLE = 'openai'
APP_ID_DALLE = 'dall-e'
MODEL_ID_DALLE = 'dall-e-3'
MODEL_VERSION_ID_DALLE = 'dc9dcb6ee67543cebc0b9a025861b868'
# Set your Clarifai credentials and model details for GPT-4 Turbo
PAT_GPT4 = 'bfdeb4029ef54d23a2e608b0aa4c00e4'
USER_ID_GPT4 = 'openai'
APP_ID_GPT4 = 'chat-completion'
MODEL_ID_GPT4 = 'gpt-4-turbo'
MODEL_VERSION_ID_GPT4 = '182136408b4b4002a920fd500839f2c8'
# Streamlit app
st.title("Pyresearch AI Integration App")
# Choose model type
model_type = st.radio("Select Model Type", ["DALL-E", "GPT-4 Turbo"])
# Input text prompt from the user
raw_text = st.text_input("Enter a text prompt:", 'ocr check mistake with image base with python opencv computer vision help out to know people')
# Button to generate result
if st.button("Generate Result"):
if model_type == "DALL-E":
# Connect to Clarifai API for DALL-E
channel_dalle = ClarifaiChannel.get_grpc_channel()
stub_dalle = service_pb2_grpc.V2Stub(channel_dalle)
metadata_dalle = (('authorization', 'Key ' + PAT_DALLE),)
userDataObject_dalle = resources_pb2.UserAppIDSet(user_id=USER_ID_DALLE, app_id=APP_ID_DALLE)
# Make a request to Clarifai API for DALL-E
post_model_outputs_response_dalle = stub_dalle.PostModelOutputs(
service_pb2.PostModelOutputsRequest(
user_app_id=userDataObject_dalle,
model_id=MODEL_ID_DALLE,
version_id=MODEL_VERSION_ID_DALLE,
inputs=[
resources_pb2.Input(
data=resources_pb2.Data(
text=resources_pb2.Text(
raw=raw_text
)
)
)
]
),
metadata=metadata_dalle
)
# Display the generated image if the request is successful
if post_model_outputs_response_dalle.status.code != status_code_pb2.SUCCESS:
st.error(f"DALL-E API request failed: {post_model_outputs_response_dalle.status.description}")
else:
output_dalle = post_model_outputs_response_dalle.outputs[0].data.image.base64
st.image(output_dalle, caption='Generated Image (DALL-E)', use_column_width=True)
elif model_type == "GPT-4 Turbo":
# Connect to Clarifai API for GPT-4 Turbo
channel_gpt4 = ClarifaiChannel.get_grpc_channel()
stub_gpt4 = service_pb2_grpc.V2Stub(channel_gpt4)
metadata_gpt4 = (('authorization', 'Key ' + PAT_GPT4),)
userDataObject_gpt4 = resources_pb2.UserAppIDSet(user_id=USER_ID_GPT4, app_id=APP_ID_GPT4)
# Make a request to Clarifai API for GPT-4 Turbo
post_model_outputs_response_gpt4 = stub_gpt4.PostModelOutputs(
service_pb2.PostModelOutputsRequest(
user_app_id=userDataObject_gpt4,
model_id=MODEL_ID_GPT4,
version_id=MODEL_VERSION_ID_GPT4,
inputs=[
resources_pb2.Input(
data=resources_pb2.Data(
text=resources_pb2.Text(
raw=raw_text
)
)
)
]
),
metadata=metadata_gpt4
)
# Display the generated result if the request is successful
if post_model_outputs_response_gpt4.status.code != status_code_pb2.SUCCESS:
st.error(f"GPT-4 Turbo API request failed: {post_model_outputs_response_gpt4.status.description}")
else:
output_gpt4 = post_model_outputs_response_gpt4.outputs[0].data.image.base64
st.image(output_gpt4, caption='Generated Image (GPT-4 Turbo)', use_column_width=True)
# Add the beautiful social media icon section
st.markdown("""
""", unsafe_allow_html=True)
