def __init__(self):
self.global_step = 1
self.max_f1 = 0
self.keras_ema = ExponentialMovingAverage(model,
decay=config.ema_decay,
temp_model=os.path.join(
config.path,
'temp_model.h5'),
type='cpu')
super(Callback, self).__init__()
class QANet_callback(Callback):
def __init__(self):
self.global_step = 1
self.max_f1 = 0
self.keras_ema = ExponentialMovingAverage(model,
decay=config.ema_decay,
temp_model=os.path.join(
config.path,
'temp_model.h5'),
type='cpu')
super(Callback, self).__init__()
def on_train_begin(self, logs=None):
lr = min(
config.learning_rate, config.learning_rate /
np.log(config.warm_up_steps) * np.log(self.global_step))
K.set_value(self.model.optimizer.lr, lr)
def on_batch_end(self, batch, logs=None):
self.global_step += 1
lr = min(
config.learning_rate, config.learning_rate /
np.log(config.warm_up_steps) * np.log(self.global_step))
K.set_value(self.model.optimizer.lr, lr)
self.keras_ema.average_update()
def on_epoch_end(self, epoch, logs=None):
self.keras_ema.assign_shadow_weights()
logits1, logits2, _, _ = self.model.predict(
x=[
dev_data['context_id'], dev_data['question_id'],
dev_data['context_char_id'], dev_data['question_char_id']
],
batch_size=config.batch_size,
verbose=1)
all_results = []
for i, qid in enumerate(dev_data['qid']):
start_logits = logits1[i, :]
end_logits = logits2[i, :]
all_results.append(
RawResult(qid=qid,
start_logits=start_logits,
end_logits=end_logits))
output_prediction_file = os.path.join(config.path,
'output_prediction.json')
output_nbest_file = os.path.join(config.path, 'output_nbest.json')
write_predictions(eval_examples,
eval_features,
all_results,
n_best_size=20,
max_answer_length=config.ans_limit,
do_lower_case=False,
output_prediction_file=output_prediction_file,
output_nbest_file=output_nbest_file)
metrics = evaluate('original_data/dev-v1.1.json',
output_prediction_file, None)
ems.append(metrics['exact'])
f1s.append(metrics['f1'])
result = pd.DataFrame([ems, f1s], index=['em', 'f1']).transpose()
result.to_csv('logs/result_' + config.name + '.csv', index=None)
if f1s[-1] > self.max_f1:
self.max_f1 = f1s[-1]
model.save_weights(
os.path.join(config.path,
'QANet_model_' + config.name + '.h5'))
model.load_weights(self.keras_ema.temp_model)