app.py

from tensorflow.keras.layers import GRU
import tensorflow as tf
from tensorflow import keras
import gradio as gr

# Define a custom GRU layer that ignores unsupported arguments
class CustomGRU(GRU):
    def __init__(self, **kwargs):
        kwargs.pop('time_major', None)
        kwargs.pop('implementation', None)
        super().__init__(**kwargs)

# Register the custom layer
custom_objects = {'CustomGRU': CustomGRU}

# Load the model with the custom objects
reconstructed_model = keras.models.load_model("poems_generation_GRU (1).h5", custom_objects=custom_objects)

def generate_text(model, temperature, start_string):
    char2idx={'\t': 0, '\n': 1, ' ': 2, 'ء': 3, 'آ': 4, 'أ': 5, 'ؤ': 6, 'إ': 7, 'ئ': 8, 'ا': 9, 'ب': 10, 'ة': 11, 'ت': 12, 'ث': 13, 'ج': 14, 'ح': 15, 'خ': 16, 'د': 17, 'ذ': 18, 'ر': 19, 'ز': 20, 'س': 21, 'ش': 22, 'ص': 23, 'ض': 24, 'ط': 25, 'ظ': 26, 'ع': 27, 'غ': 28, 'ف': 29, 'ق': 30, 'ك': 31, 'ل': 32, 'م': 33, 'ن': 34, 'ه': 35, 'و': 36, 'ى': 37, 'ي': 38}
    idx2char=['\t', '\n', ' ', 'ء', 'آ', 'أ', 'ؤ', 'إ', 'ئ', 'ا', 'ب', 'ة', 'ت', 'ث', 'ج', 'ح', 'خ', 'د', 'ذ', 'ر', 'ز', 'س', 'ش', 'ص', 'ض', 'ط', 'ظ', 'ع', 'غ', 'ف', 'ق', 'ك', 'ل', 'م', 'ن', 'ه', 'و', 'ى', 'ي']
    # Evaluation step (generating text using the learned model)

    # Number of characters to generate
    num_generate = 1000

    # Converting our start string to numbers (vectorizing)
    input_eval = [char2idx[s] for s in start_string]
    input_eval = tf.expand_dims(input_eval, 0)

    # Empty string to store our results
    text_generated = []

    # Low temperatures results in more predictable text.
    # Higher temperatures results in more surprising text.
    # Experiment to find the best setting.

    # Here batch size == 1
    model.reset_states()
    for i in range(num_generate):
        predictions = model(input_eval)
        # remove the batch dimension
        predictions = tf.squeeze(predictions, 0)

        # using a random.categorical distribution to predict the word returned by the model
        predictions = predictions / temperature
        predicted_id = tf.random.categorical(predictions, num_samples=1)[-1,0].numpy()
        
        input_eval = tf.expand_dims([predicted_id], 0)

        text_generated.append(idx2char[predicted_id])

    return (start_string + ''.join(text_generated))

def generate_poem(start, temperature):
    return generate_text(reconstructed_model, temperature, start_string=u""+start)

iface = gr.Interface(fn=generate_poem, inputs=["text", gr.Slider(0, 1, value=1)], outputs="text")
iface.launch()