Create new file

app.py

@@ -0,0 +1,55 @@
+import tensorflow as tf 
+import tensorflow_text as text
+from tensorflow.train import Checkpoint
+import pandas as pd
+import numpy as np
+import gradio
+from Model import Transformer
+
+vocab = []
+with open("vocab.txt", mode = "r", encoding = "utf-8") as file:
+    for token in file:
+        vocab.append(token.replace("\n", ""))
+        
+tokenizer = text.FastBertTokenizer(vocab, support_detokenization = True)
+
+VOCAB_SIZE = len(vocab)
+D_MODEL = 256
+NB_LAYERS = 6
+FFN_UNITS = 2048
+NB_PROJ = 8
+DROPOUT_RATE = 0.1
+
+transformer = Transformer(vocab_size_enc = VOCAB_SIZE, 
+                          vocab_size_dec = 1,
+                          d_model = D_MODEL,
+                          nb_layers = NB_LAYERS,
+                          FFN_units = FFN_UNITS,
+                          nb_proj = NB_PROJ,
+                          dropout_rate = DROPOUT_RATE)
+                          
+ckpt = Checkpoint()   
+ckpt.restore("ckpt-10")  
+print("Checkpoint Restaurado")  
+
+def evaluate(sentence):
+    ragged = tokenizer.tokenize([sentence])
+    ragged = trimer.trim([ragged])[0]
+    count = ragged.bounding_shape()[0]
+    starts = tf.fill([count,1], START)
+    ends = tf.fill([count,1], END)
+    inputs = tf.concat([starts, ragged, ends], axis=1)
+    inputs, _ = text.pad_model_inputs(inputs, max_seq_length = MAX_LENGTH + 2, pad_value = PAD)
+    
+    prediction = transformer(inputs, False)
+    
+    prediction = tf.round(prediction)
+    if prediction == 0:
+        return"Negative"
+    else:
+        return"Positive"
+        
+app = gr.Interface(fn = evaluate, title = "IMDb Sentiment Classifier", description = "Write a sentence with a positive or negative sentiment", inputs = "text", outputs = "text")   
+app.launch(share = True)                                                                                  
+
+