Add STS-B demo

demo/GLUE/STS-B/stsb.py

@@ -0,0 +1,110 @@
+import os
+
+import numpy as np
+from scipy.stats import pearsonr, spearmanr
+
+from keras_xlnet.backend import keras
+from keras_bert.layers import Extract
+from keras_xlnet import PretrainedList, get_pretrained_paths
+from keras_xlnet import Tokenizer, load_trained_model_from_checkpoint, ATTENTION_TYPE_BI
+
+EPOCH = 10
+BATCH_SIZE = 32
+SEQ_LEN = 140
+MODEL_NAME = 'STS-B.h5'
+
+current_path = os.path.dirname(os.path.abspath(__file__))
+train_path = os.path.join(current_path, 'train.tsv')
+dev_path = os.path.join(current_path, 'dev.tsv')
+
+paths = get_pretrained_paths(PretrainedList.en_cased_base)
+tokenizer = Tokenizer(paths.vocab)
+
+
+# Read data
+class DataSequence(keras.utils.Sequence):
+
+    def __init__(self, x, y):
+        self.x = x
+        self.y = y
+
+    def __len__(self):
+        return (len(self.y) + BATCH_SIZE - 1) // BATCH_SIZE
+
+    def __getitem__(self, index):
+        s = slice(index * BATCH_SIZE, (index + 1) * BATCH_SIZE)
+        return [item[s] for item in self.x], self.y[s]
+
+
+def generate_sequence(path):
+    tokens, classes, scores = [], [], []
+    max_len = 0
+    with open(path) as reader:
+        reader.readline()
+        for line in reader:
+            line = line.strip()
+            parts = line.split('\t')
+            encoded_a, encoded_b = tokenizer.encode(parts[7]), tokenizer.encode(parts[8])
+            encoded = encoded_a + [tokenizer.SYM_SEP] + encoded_b + [tokenizer.SYM_SEP]
+            max_len = max(max_len, len(encoded))
+            encoded = [tokenizer.SYM_PAD] * (SEQ_LEN - 1 - len(encoded)) + encoded + [tokenizer.SYM_CLS]
+            tokens.append(encoded)
+            classes.append(round(float(parts[9])))
+            scores.append(float(parts[9]))
+    tokens, classes = np.array(tokens), np.array(classes)
+    segments = np.zeros_like(tokens)
+    segments[:, -1] = 1
+    lengths = np.zeros_like(tokens[:, :1])
+    return DataSequence([tokens, segments, lengths], classes), scores
+
+
+current_path = os.path.dirname(os.path.abspath(__file__))
+train_seq, _ = generate_sequence(train_path)
+dev_seq, scores = generate_sequence(dev_path)
+
+# Load pretrained model
+model = load_trained_model_from_checkpoint(
+    config_path=paths.config,
+    checkpoint_path=paths.model,
+    batch_size=BATCH_SIZE,
+    memory_len=0,
+    target_len=SEQ_LEN,
+    in_train_phase=False,
+    attention_type=ATTENTION_TYPE_BI,
+)
+
+
+# Build classification model
+last = Extract(index=-1, name='Extract')(model.output)
+dense = keras.layers.Dense(units=768, activation='tanh', name='Dense')(last)
+dropout = keras.layers.Dropout(rate=0.1, name='Dropout')(dense)
+output = keras.layers.Dense(units=6, activation='softmax', name='Softmax')(dropout)
+model = keras.models.Model(inputs=model.inputs, outputs=output)
+model.summary()
+
+
+# Fit model
+if os.path.exists(MODEL_NAME):
+    model.load_weights(MODEL_NAME)
+
+model.compile(
+    optimizer=keras.optimizers.Adam(lr=3e-5),
+    loss='sparse_categorical_crossentropy',
+    metrics=['sparse_categorical_accuracy'],
+)
+
+model.fit_generator(
+    generator=train_seq,
+    validation_data=dev_seq,
+    epochs=EPOCH,
+    callbacks=[keras.callbacks.EarlyStopping(monitor='val_loss', patience=2)],
+)
+
+model.save_weights(MODEL_NAME)
+
+# Evaluation
+# Use dev set because the results of test set is unknown
+classes = np.array([[0], [1], [2], [3], [4], [5]])
+results = np.dot(model.predict_generator(dev_seq, verbose=True), classes).squeeze(axis=-1)
+print('Pearson: %.2f' % (100.0 * pearsonr(results, scores)[0]))
+print('Spearman: %.2f' % (100.0 * spearmanr(results, scores)[0]))