imagine I have the following dataset:
import pandas as pd
# Positive and negative sentences
positive_sentences = [
"I love this product!",
"The weather is beautiful today.",
"The team did an excellent job.",
"She is a very talented musician."
]
negative_sentences = [
"I am not satisfied with the service.",
"The food was terrible at that restaurant.",
"The movie was a complete disappointment.",
"He made a lot of mistakes in the project."
]
# Combine positive and negative sentences
sentences = positive_sentences + negative_sentences
# Create a DataFrame with a "snippet" column
df = pd.DataFrame({'snippet': sentences})
# Display the DataFrame
print(df)
I want to use a LLM model that answers the following question. Is the following sentence positive, negative or neutral?
This is what I have tried so far:
## installing the libraries:
import pandas as pd
from transformers import pipeline, AutoModelForQuestionAnswering, AutoTokenizer
from transformers import RobertaTokenizer
from transformers import AutoTokenizer, RobertaForQuestionAnswering
import torch
from transformers import BertForQuestionAnswering, BertTokenizer
# Setting up the model and tokenizer
tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
model = BertForQuestionAnswering.from_pretrained('bert-large-uncased-whole-word-masking-finetuned-squad')
# Create a question answering pipeline
question_answerer = pipeline("question-answering", model=model, tokenizer=tokenizer)
for index, row in df.iterrows():
article = row["snippet"]
prompt = f"Is the following sentence positive, negative or neutral? {article}"
result = question_answerer(question=prompt, context=article)
# Check if "answer" key is in the result
if "answer" in result:
main_theme = result["answer"]
print(f"Article {index+1} main theme: {main_theme}")
else:
print(f"Article {index+1} main theme not found in the result.")