Spaces:
Sleeping
Sleeping
| # AUTOGENERATED! DO NOT EDIT! File to edit: ../drive/MyDrive/Codici/Python/Apps/Gradio_App/SemanticSearch_QA-v2.1.ipynb. | |
| # %% auto 0 | |
| __all__ = ['model_name', 'qa_model', 'contexts', 'question', 'df_results', 'question_1', 'question_2', 'question_3', 'question_4', | |
| 'question_5', 'question_6', 'question_7', 'question_8', 'question_9', 'question_10', 'title', 'description', | |
| 'data', 'context_df', 'question_input', 'contexts_input', 'n_answers_input', 'full_context_input', | |
| 'confidence_threshold_input', 'intf', 'get_answers'] | |
| # %% ../drive/MyDrive/Codici/Python/Apps/Gradio_App/SemanticSearch_QA-v2.1.ipynb 3 | |
| import pandas as pd | |
| import gradio as gr | |
| import transformers | |
| from transformers import AutoModelForQuestionAnswering, AutoTokenizer, pipeline | |
| # model_name = 'Francesco-A/bert-finetuned-squad-v1' | |
| model_name = "deepset/roberta-base-squad2" | |
| qa_model = pipeline(task = 'question-answering', | |
| model = model_name, | |
| tokenizer = model_name) | |
| # %% ../drive/MyDrive/Codici/Python/Apps/Gradio_App/SemanticSearch_QA-v2.1.ipynb 6 | |
| def get_answers(question, contexts, n_answers=1, full_context=True, confidence_threshold = 0.5): | |
| results = [] | |
| if isinstance(contexts, pd.DataFrame): | |
| # If it's a DataFrame, get the values from the 'Context' column as a list | |
| contexts = contexts['Context'].to_list() | |
| for i, context in enumerate(contexts): | |
| QA_input = {'question': question, 'context': context} | |
| res = qa_model(question = QA_input['question'], context = QA_input['context']) | |
| results_dict = { | |
| 'context_idx': i, | |
| # 'Question': question, | |
| 'Answer': res['answer'], | |
| 'Score': round(res['score'],3) | |
| } | |
| if full_context: | |
| results_dict['Full Context'] = context | |
| results.append(results_dict) | |
| df = pd.DataFrame(results) | |
| df = df[df['Score'] >= confidence_threshold] | |
| df = df.sort_values(by='Score', ascending=False).head(n_answers) | |
| df = df.reset_index(drop=True) # Reset index after sorting | |
| return df | |
| # Example usage: | |
| contexts = [ | |
| 'The option to convert models between FARM and transformers gives freedom to the user and let people easily switch between frameworks.', | |
| 'Model conversion enables interoperability between different NLP libraries.', | |
| 'Converting models allows for leveraging the strengths of various tools.' | |
| ] | |
| question = "Why is model conversion important?" | |
| df_results = get_answers(question,contexts,n_answers=2,full_context=False, confidence_threshold = 0.25) | |
| df_results | |
| # %% ../drive/MyDrive/Codici/Python/Apps/Gradio_App/SemanticSearch_QA-v2.1.ipynb 7 | |
| # Define example question(s) | |
| question_1 = "What are the main features of the new XPhone 20?" | |
| question_2 = "What are some benefits of regular exercise?" | |
| question_3 = "What is the color of a rose?" | |
| question_4 = "What's photosynthesis?" | |
| question_5 = "At what temperature does water boil?" | |
| question_6 = "Where can I find potassium?" | |
| question_7 = "How does the internet function?" | |
| question_8 = "What are the ingredients for making a classic margarita?" | |
| question_9 = "How does cellular respiration work?" | |
| question_10 = "Is money important?" | |
| # Define example contexts as a list of strings | |
| contexts = [ | |
| "The XPhone 20 is expected to come with an improved camera system, featuring advanced image stabilization and enhanced low-light capabilities.", | |
| "Regular exercise has been shown to reduce the risk of chronic diseases such as heart disease, diabetes, and certain types of cancer.", | |
| "Roses come in various colors, including red, pink, yellow, white, and even blue (though blue roses are rare and often created through genetic modification).", | |
| "Photosynthesis occurs in the chloroplasts of plant cells, where chlorophyll captures sunlight and converts it into chemical energy.", | |
| "Water boils at different temperatures depending on factors like altitude and atmospheric pressure. At sea level, it boils at 100 degrees Celsius or 212 degrees Fahrenheit.", | |
| "Potassium is an essential mineral that can be found in various foods such as bananas, potatoes, spinach, and oranges.", | |
| "The internet functions through a complex system of data transmission protocols, routers, and servers that allow for the exchange of information globally.", | |
| "A classic margarita typically consists of tequila, lime juice, triple sec (or orange liqueur), and is often served with a salted rim.", | |
| "Cellular respiration takes place in the mitochondria of cells, where glucose and oxygen are converted into ATP (adenosine triphosphate) and carbon dioxide.", | |
| "Money is a medium of exchange that facilitates transactions of goods and services. Its importance lies in its role in economic systems and the ability to represent value.", | |
| "The XPhone 20 may feature an OLED display for vibrant colors and deep blacks, providing a high-quality visual experience.", | |
| "Exercise releases endorphins, which are chemicals in the brain that help improve mood and reduce feelings of stress and anxiety.", | |
| "Different species of roses can have variations in color, including shades of red, pink, yellow, and white.", | |
| "During photosynthesis, plants also release oxygen as a byproduct, which is essential for the survival of many organisms on Earth.", | |
| "Water boils at a lower temperature at higher altitudes due to the reduced atmospheric pressure. For example, in the mountains, it may boil below 100 degrees Celsius.", | |
| "Potassium is crucial for proper muscle function, nerve function, and maintaining fluid balance in the body.", | |
| "The internet relies on a system of interconnected networks, including wired and wireless connections, to transmit data across the globe.", | |
| "In addition to the core ingredients, a classic margarita can also be garnished with a wedge of lime for added flavor.", | |
| "Cellular respiration involves several stages, including glycolysis, the Krebs cycle, and the electron transport chain, to extract energy from glucose.", | |
| "Money serves as a unit of account, allowing for standardized pricing and valuation of goods and services in economies worldwide." | |
| "The XPhone 20 is rumored to feature a smaller notch, providing more screen real estate for users. This allows for an immersive viewing experience.", | |
| "Photosynthesis is the process by which plants convert carbon dioxide, water, and sunlight into glucose and oxygen.", | |
| "Bananas are a great source of potassium.", | |
| "The theory of relativity was developed by Albert Einstein and revolutionized our understanding of space and time.", | |
| "The Eiffel Tower is located in Paris, France.", | |
| "Reports suggest that the XPhone 20 will have significant improvements in battery life compared to its predecessor. Users can expect a longer-lasting device.", | |
| "A penny saved is a penny earned.", | |
| "Water boils at 100 degrees Celsius.", | |
| "The Great Wall of China is one of the most impressive architectural feats in history.", | |
| "The capital of Japan is Tokyo.", | |
| "One of the anticipated features of the XPhone 20 is a faster and more powerful A16 chip. This will result in smoother and more efficient performance.", | |
| "Roses are red, violets are blue.", | |
| "Regular exercise can help improve cardiovascular health and strengthen muscles.", | |
| "A classic margarita is made with tequila, lime juice, and orange liqueur.", | |
| "Cellular respiration is the process by which cells convert glucose and oxygen into energy, carbon dioxide, and water.", | |
| "The internet is a global network of interconnected computers and servers that allows the sharing of information and resources.", | |
| "Mount Everest is the highest mountain in the world, located in the Himalayas.", | |
| "The sun rises in the east and sets in the west.", | |
| "The Mona Lisa is a famous portrait painting by Leonardo da Vinci.", | |
| "The Declaration of Independence was adopted by the Continental Congress on July 4, 1776.", | |
| ] | |
| # %% ../drive/MyDrive/Codici/Python/Apps/Gradio_App/SemanticSearch_QA-v2.1.ipynb 10 | |
| title = 'SemanticSearch_QA-v2' | |
| description = """ | |
| QA model: [deepset/roberta-base-squad2](https://huggingface.co/deepset/roberta-base-squad2) | |
| """ | |
| data = { | |
| 'Context': contexts, | |
| } | |
| context_df = pd.DataFrame(data) | |
| question_input = gr.Textbox(label="Question", placeholder="Enter your question here") | |
| contexts_input = gr.Dataframe(value=context_df, headers=["Context"], interactive=True, type="pandas", label="Contexts") | |
| n_answers_input = gr.Slider(minimum=1, maximum=10, step=1, value=1, label="Number of Answers") | |
| full_context_input = gr.Checkbox(label="Include Full Context", value=True) | |
| confidence_threshold_input = gr.Slider(minimum=0.0, maximum=1.0, step=0.1, value=0.5, label="Confidence Threshold") | |
| intf = gr.Interface(fn=get_answers, | |
| inputs= [question_input, contexts_input, n_answers_input,full_context_input,confidence_threshold_input], | |
| outputs= gr.components.Dataframe(type="pandas", row_count=5), | |
| examples = [[question_1,context_df,3,False,0.1], | |
| [question_2,context_df,5,True,0.1], | |
| [question_4,context_df,10,False,0.1]], | |
| title=title, | |
| description=description, | |
| # article=long_desc | |
| ) | |
| intf.launch(inline=True, | |
| # share=True | |
| ) | |