ChatGPT Model Selection – Dr. Trippers Stuff

I need a streamlit app that mimics the chatgpt user interface. In place of the options for gpt 3.5 and gpt 4 use Pipeline and Local for the labels and then code the rest of the document around this: # Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained(“Open-Orca/oo-phi-1_5”, trust_remote_code=True)
I need a streamlit app that mimics the chatgpt user interface. In place of the options for gpt 3.5 and gpt 4 use Pipeline and Local for the labels and then code the rest of the document around this:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline(“text-generation”, model=”Open-Orca/oo-phi-1_5″, trust_remote_code=True)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained(“Open-Orca/oo-phi-1_5”, trust_remote_code=True)
To create a Streamlit app that mimics the ChatGPT user interface using the specified model and labels, you can follow these steps. First, make sure you have Streamlit and Transformers libraries installed. You can install them using pip if you haven’t already:

“`bash
pip install streamlit transformers
“`

Next, create a Python script (e.g., `chatgpt_app.py`) with the following code:

“`python
import streamlit as st
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline

# Load model directly
model = AutoModelForCausalLM.from_pretrained(“Open-Orca/oo-phi-1_5”, trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained(“Open-Orca/oo-phi-1_5”)

# Create a Streamlit app
st.title(“ChatGPT Mimic”)

# Add a sidebar for options
st.sidebar.header(“Options”)
selected_option = st.sidebar.radio(“Select Option”, [“Pipeline”, “Local”])

# Initialize the ChatGPT pipeline or local model
if selected_option == “Pipeline”:
chatgpt = pipeline(“text-generation”, model=model, tokenizer=tokenizer)
else:
def generate_response(input_text, max_length=50):
input_ids = tokenizer.encode(input_text, return_tensors=”pt”)
output = model.generate(input_ids, max_length=max_length, num_return_sequences=1)
response = tokenizer.decode(output[0], skip_special_tokens=True)
return response

# Create a text input field for user input
user_input = st.text_input(“You:”, “”)

# Define a function to
To create a Streamlit app that mimics the ChatGPT user interface and replaces the options for GPT 3.5 and GPT 4 with “Pipeline” and “Local” labels, you can follow these steps. I’ll provide you with a concise yet complete code example.

First, make sure you have Streamlit and Transformers libraries installed. You can install them using pip:

“`bash
pip install streamlit transformers
“`

Next, create a Python script (e.g., `chatgpt_app.py`) with the following code:

“`python
import streamlit as st
from transformers import pipeline, AutoModelForCausalLM

# Title
st.title(“ChatGPT Mimic”)

# Sidebar options
model_option = st.sidebar.selectbox(“Select Model”, [“Pipeline”, “Local”])

# Load the model based on the selected option
if model_option == “Pipeline”:
# Use a pipeline as a high-level helper
pipe = pipeline(“text-generation”, model=”Open-Orca/oo-phi-1_5″, trust_remote_code=True)
else:
# Load model directly
model = AutoModelForCausalLM.from_pretrained(“Open-Orca/oo-phi-1_5”, trust_remote_code=True)

# User input
user_input = st.text_input(“You:”, “”)

if user_input:
# Generate a response
if model_option == “Pipeline”:
response = pipe(user_input, max_length=50, num_return_sequences=1, do_sample=True)
st.write(“ChatGPT:”, response[0][‘generated_text’])
else:
st.write(“ChatGPT:”, “Sorry, local model execution is not supported in this demo.”)

# Instructions
st.sidebar.markdown(“### Instructions”)
st.sidebar.write(“1. Select a model from the sidebar.”)
st.sidebar.write(“2. Type your message in the text input box.”)
st.sidebar.write(“3. Press Enter to get a response.”)

# Footer
st.sidebar.markdown(“—“)
st.sidebar.write(“This Streamlit app mimics the ChatGPT user interface.”)
st.sidebar.write(“Replace ‘Pipeline’ and ‘Local’ with your model names and code accordingly.”)

# Display the Streamlit app
if __name__ == “__main__”:
st.sidebar.markdown(“—“)
st.sidebar.write(“© 2023 Your Company”)
“`

In this script:

1. We use Streamlit to create a simple web app.
2. The user can select a model from the sidebar, either “Pipeline” or “Local.”
3. If “Pipeline” is selected, we use the Hugging Face Transformers library’s pipeline to generate responses.
4. If “Local” is selected, we load the model directly, but this part is a placeholder since local model execution is not supported in this example.
5. The user can enter their text input, and the app will generate a response accordingly.
6. Instructions and a footer are provided in the sidebar.

Make sure to replace “Pipeline” and “Local” with your actual model names and adjust the code accordingly.

To run the app, use the following command:

“`bash
streamlit run chatgpt_app.py
“`

This will launch a web app where you can interact with the ChatGPT-like interface using the selected model.
this will run the pipeline slection though
I apologize for the confusion. If you want to run either the pipeline or load the model directly based on the user’s selection, you can modify the code as follows:

“`python
import streamlit as st
from transformers import pipeline, AutoModelForCausalLM

# Title
st.title(“ChatGPT Mimic”)

# Sidebar options
model_option = st.sidebar.radio(“Select Model”, [“Pipeline”, “Local”])

# User input
user_input = st.text_input(“You:”, “”)

# Load the model based on the selected option
if model_option == “Pipeline”:
# Use a pipeline as a high-level helper
chat_model = pipeline(“text-generation”, model=”Open-Orca/oo-phi-1_5″, trust_remote_code=True)
else:
# Load model directly
chat_model = AutoModelForCausalLM.from_pretrained(“Open-Orca/oo-phi-1_5”, trust_remote_code=True)

# Function to generate a response
def generate_response(input_text, model):
if model_option == “Pipeline”:
response = model(input_text, max_length=50, num_return_sequences=1, do_sample=True)
return response[0][‘generated_text’]
else:
return “Sorry, local model execution is not supported in this demo.”

if user_input:
response = generate_response(user_input, chat_model)
st.write(“ChatGPT:”, response)

# Display the Streamlit app
if __name__ == “__main__”:
st.sidebar.markdown(“—“)
st.sidebar.write(“© 2023 Your Company”)
“`

Now, the model selection will determine whether to use the pipeline or load the model directly. This code should work as expected, running the selected model based on user input.