| import gradio as gr | |
| import torch | |
| from transformers import XLNetTokenizer, XLNetModel | |
| import numpy as np | |
| import joblib | |
| device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') | |
| tokenizer = XLNetTokenizer.from_pretrained('xlnet-base-cased') | |
| xlnet_model = XLNetModel.from_pretrained('xlnet-base-cased').to(device) | |
| random_forest_classifier = joblib.load("random_forest_model.pkl") | |
| def get_embedding(text): | |
| inputs = tokenizer(text, return_tensors='pt', truncation=True, padding=True, max_length=512).to(device) | |
| with torch.no_grad(): | |
| outputs = xlnet_model(**inputs) | |
| return outputs.last_hidden_state.mean(dim=1).squeeze().cpu().numpy() | |
| def predict_complexity(sentence, target_word): | |
| try: | |
| sentence_embedding = get_embedding(sentence) | |
| word_embedding = get_embedding(target_word) | |
| combined_embedding = np.concatenate([sentence_embedding, word_embedding]).reshape(1, -1) | |
| prediction = random_forest_classifier.predict(combined_embedding)[0] | |
| return f"🔍 Predicted Complexity Level: **{prediction}**" | |
| except Exception as e: | |
| return f"❌ Error: {str(e)}" | |
| with gr.Blocks() as demo: | |
| gr.Markdown("## ✨ Word Complexity Predictor") | |
| with gr.Row(): | |
| sentence_input = gr.Textbox(label="Full Sentence", placeholder="Type a full sentence...") | |
| word_input = gr.Textbox(label="Target Word", placeholder="Type the target word...") | |
| output = gr.Markdown() | |
| gr.Button("Predict Complexity").click(predict_complexity, [sentence_input, word_input], output) | |
| demo.launch() | |