File size: 1,259 Bytes
58eb302 2d2d40c 58eb302 2d2d40c 2af70a4 9bdeb47 2d2d40c 9bdeb47 2d2d40c 9bdeb47 2d2d40c |
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 |
import gradio as gr
from langchain import HuggingFacePipeline, PromptTemplate, LLMChain
from transformers import AutoTokenizer
import transformers
import torch
# Define the Hugging Face model
model = "manohar02/Llama-2-7b-finetune"
# Define the Hugging Face pipeline
pipeline = transformers.pipeline(
"text-generation", # task
model=model,
torch_dtype=torch.bfloat16,
trust_remote_code=True,
device_map="auto",
max_length=20000,
do_sample=True,
top_k=10,
num_return_sequences=1,
eos_token_id=AutoTokenizer.from_pretrained(model).eos_token_id
)
llm = HuggingFacePipeline(pipeline=pipeline, model_kwargs={'temperature': 0})
# Define the template for summarization
template = """
Write a concise summary of the following text delimited by triple backquotes.
'''{text}'''
SUMMARY:
"""
prompt = PromptTemplate(template=template, input_variables=["text"])
llm_chain = LLMChain(prompt=prompt, llm=llm)
# Function to generate summary
def generate_summary(text):
summary = llm_chain.run(text)
# Extract only the summary part from the output
return summary.strip()
# Create a Gradio interface
iface = gr.Interface(fn=generate_summary, inputs="text", outputs="text", title="LLaMA2 Summarizer")
iface.launch()
|