Spaces:
Runtime error
Runtime error
File size: 2,552 Bytes
5aa525a 1d7f870 5aa525a 587858f 5aa525a 0f3d4bc af29fad 5aa525a b65da23 5aa525a 92cca2a 1d7f870 5aa525a af29fad 5aa525a 7d18afe af96bbe af29fad 5aa525a 3952e87 af29fad 5aa525a 898c8a2 ed78754 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 |
"""
@author: idoia lerchundi
"""
import os
import streamlit as st
from huggingface_hub import InferenceClient
import random
# Load the API token from an environment variable
api_key = os.getenv("HF_TOKEN")
# Instantiate the InferenceClient
client = InferenceClient(api_key=api_key)
# Streamlit app title
st.title("Text-generation model using Streamlit from Inference API (serverless) feature.")
# Ensure the full_text key is initialized in session state
if "full_text" not in st.session_state:
st.session_state["full_text"] = ""
# Model selection dropdown
model_options = [ "TinyLlama/TinyLlama-1.1B-Chat-v1.0", "gpt2", "facebook/opt-1.3b", "EleutherAI/gpt-neo-2.7B","meta-llama/Llama-Llama-3-8B-Instruct", "meta-llama/Llama-Llama-3.1-1B-Instruct", "meta-llama/Llama-Llama-3.2-3B-Instruct", "meta-llama/Llama-Llama-3.2-8B-Instruct", "Qwen/Qwen2.5-1.5B-Instruct", "openai-community/gpt2", "google/gemma-1.1-7b-it", "google/gemma-1.27b-it", "google/gemma-1.2b-it", "google/gemma-1.9b-it", "google/gemma-2.2b-it", "HuggingFaceH4/starchat7b-beta", "distilbert/distilgpt2", "facebook/opt-1.3b", "distributed/optimized=gpt2-1b" ]
selected_model = st.selectbox("Choose a model:"= st.selectbox("Choose a model:", model_options)
# Create a text input area for user prompts
with st.form("my_form"):
text = st.text_area("JOKER (TinyLlama is not great at joke telling.) (using model TinyLlama/TinyLlama-1.1B-Chat-v1.0):", "Tell me a clever and funny joke in exactly 4 sentences. It should make me laugh really hard. Don't repeat the topic in your joke. Be creative and concise.")
submitted = st.form_submit_button("Submit")
# Initialize the full_text variable
full_text = " "
# Generate a random temperature between 0.5 and 1.0
temperature = random.uniform(0.5, 1.0)
if submitted:
messages = [
{"role": "user", "content": text}
]
# Create a new stream for each submission
stream = client.chat.completions.create(
model=selected_model,
messages=messages,
# Generate a random temperature between 0.5 and 1.0
temperature = random.uniform(0.5, 1.0),
max_tokens=300,
top_p=random.uniform(0.7, 1.0),
stream=True
)
# Concatenate chunks to form the full response
for chunk in stream:
full_text += chunk.choices[0].delta.content
# Update session state with the full response
st.session_state["full_text"] = full_text
# Display the full response
if st.session_state["full_text"]:
st.info(st.session_state["full_text"]) |