app.py

from huggingface_hub import from_pretrained_keras
import numpy as np
import gradio as gr
import transformers
import tensorflow as tf

class BertSemanticDataGenerator(tf.keras.utils.Sequence):
    """Generates batches of data."""
    def __init__(
        self,
        sentence_pairs,
        labels,
        batch_size=32,
        shuffle=True,
        include_targets=True,
    ):
        self.sentence_pairs = sentence_pairs
        self.labels = labels
        self.shuffle = shuffle
        self.batch_size = batch_size
        self.include_targets = include_targets
        # Load our BERT Tokenizer to encode the text.
        # We will use base-base-uncased pretrained model.
        self.tokenizer = transformers.BertTokenizer.from_pretrained(
            "bert-base-uncased", do_lower_case=True
        )
        self.indexes = np.arange(len(self.sentence_pairs))
        self.on_epoch_end()

    def __len__(self):
        # Denotes the number of batches per epoch.
        return len(self.sentence_pairs) // self.batch_size

    def __getitem__(self, idx):
        # Retrieves the batch of index.
        indexes = self.indexes[idx * self.batch_size : (idx + 1) * self.batch_size]
        sentence_pairs = self.sentence_pairs[indexes]

        # With BERT tokenizer's batch_encode_plus batch of both the sentences are
        # encoded together and separated by [SEP] token.
        encoded = self.tokenizer.batch_encode_plus(
            sentence_pairs.tolist(),
            add_special_tokens=True,
            max_length=128,
            return_attention_mask=True,
            return_token_type_ids=True,
            pad_to_max_length=True,
            return_tensors="tf",
        )

        # Convert batch of encoded features to numpy array.
        input_ids = np.array(encoded["input_ids"], dtype="int32")
        attention_masks = np.array(encoded["attention_mask"], dtype="int32")
        token_type_ids = np.array(encoded["token_type_ids"], dtype="int32")

        # Set to true if data generator is used for training/validation.
        if self.include_targets:
            labels = np.array(self.labels[indexes], dtype="int32")
            return [input_ids, attention_masks, token_type_ids], labels
        else:
            return [input_ids, attention_masks, token_type_ids]

model = from_pretrained_keras("keras-io/bert-semantic-similarity")
labels = ["contradiction", "entailment", "neutral"]

def predict(sentence1, sentence2):
    sentence_pairs = np.array([[str(sentence1), str(sentence2)]])
    test_data = BertSemanticDataGenerator(
        sentence_pairs, labels=None, batch_size=1, shuffle=False, include_targets=False,
    )
    probs = model.predict(test_data[0])[0]
    
    labels_probs = {labels[i]: float(probs[i]) for i, _ in enumerate(labels)}
    return labels_probs
    
    #idx = np.argmax(proba)
    #proba = f"{proba[idx]*100:.2f}%"
    #pred = labels[idx]
    #return f'The semantic similarity of two input sentences is {pred} with {proba} of probability'

inputs = [
         gr.Audio(source = "upload", label='Upload audio file', type="filepath"),
]

examples = [["Two women are observing something together.", "Two women are standing with their eyes closed."],
            ["A smiling costumed woman is holding an umbrella", "A happy woman in a fairy costume holds an umbrella"],
            ["A soccer game with multiple males playing", "Some men are playing a sport"],            
]

gr.Interface(
    fn=predict,
    title="Semantic Similarity with BERT",
    description = "Natural Language Inference by fine-tuning BERT model on SNLI Corpus 📰",
    inputs=["text", "text"],
    examples=examples,
    #outputs=gr.Textbox(label='Prediction'),
    outputs=gr.outputs.Label(num_top_classes=3, label='Semantic similarity'),
    cache_examples=True,
    article = "Author: <a href=\"https://huggingface.co/vumichien\">Vu Minh Chien</a>. Based on the keras example from <a href=\"https://keras.io/examples/nlp/semantic_similarity_with_bert/\">Mohamad Merchant</a>",
).launch(debug=True, enable_queue=True)