Spaces:
Sleeping
Sleeping
File size: 2,297 Bytes
ce6e64e f1f5f03 06aa116 a187c2a f1f5f03 a187c2a f1f5f03 3792e01 6133325 3792e01 6133325 3792e01 c1367b5 06aa116 3792e01 f1f5f03 |
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 65 66 67 |
import os
import streamlit as st
from transformers import pipeline
import warnings
from huggingface_hub import login
import torch
# Title for the web app
st.title("Test Large Language Models from Hugging Face")
# Dropdown for model selection
models = [
"meta-llama/Llama-3.2-3B-Instruct",
"mistralai/Mistral-7B-Instruct-v0.3",
"ibm-granite/granite-3.0-8b-instruct",
"bartowski/Meta-Llama-3.1-8B-Claude-GGUF"
]
st.write(
#"Has environment variables been set:",
os.environ["HF_ACCESS_TOKEN"] == st.secrets["HF_ACCESS_TOKEN"])
hf_api_token = st.secrets["HF_ACCESS_TOKEN"]
login(token=hf_api_token)
selected_model = st.selectbox("Select a model to test:", models)
# Text input for user prompt
user_input = st.text_area("Enter your text prompt:", "Type something here...")
# Load the selected model using the pipeline
@st.cache_resource
#def load_model(model_name):
# return pipeline("text-generation", model=model_name, framework="tf")
# with warnings.catch_warnings():
# warnings.simplefilter("ignore")
# access_token = os.getenv("HF_ACCESS_TOKEN") # Use an environment variable for the access token
# return pipeline("text-generation", model=model_name, framework="tf", use_auth_token=access_token) # Use TensorFlow framework
def load_model(model_name):
with warnings.catch_warnings():
warnings.simplefilter("ignore")
access_token = os.getenv("HF_ACCESS_TOKEN") # Use an environment variable for the access token
return pipeline("text-generation", model=model_name, token=access_token) # Use TensorFlow framework
# Button to run the model
if st.button("Generate Response"):
if user_input:
try:
# Load the selected model
generator = load_model(selected_model)
# Generate text based on the input with truncation enabled
with st.spinner("Generating response..."):
result = generator(user_input, max_length=100, num_return_sequences=1, truncation=True)
# Display the result
st.subheader("Generated Response:")
st.write(result[0]['generated_text'])
except Exception as e:
st.error(f"An error occurred: {e}")
else:
st.warning("Please enter a text prompt.") |