def predict(args):
"""
predicts answers for test files
"""
logger = logging.getLogger("brc")
logger.info('Load data_set and vocab...')
# with open(os.path.join(args.vocab_dir, 'vocab.data'), 'rb') as fin:
with open(args.vocab_path, 'rb') as fin:
vocab = pickle.load(fin)
assert len(args.test_files) > 0, 'No test files are provided.'
brc_data = BRCDataset(args.max_p_num,
args.max_p_len,
args.max_q_len,
test_files=args.test_files)
logger.info('Converting text into ids...')
brc_data.convert_to_ids(vocab)
logger.info('Restoring the model...')
rc_model = RCModel(vocab, args)
rc_model.restore(model_dir=args.model_dir, model_prefix=args.algo)
logger.info('Predicting answers for test set...')
test_batches = brc_data.gen_mini_batches('test',
args.batch_size,
pad_id=vocab.get_id(
vocab.pad_token),
shuffle=False)
rc_model.evaluate(test_batches,
result_dir=args.result_dir,
result_prefix='test.predicted')
def predict(args):
"""
预测测试文件的答案
"""
logger = logging.getLogger("brc")
logger.info('加载数据集和词汇表...')
with open(os.path.join(args.vocab_dir, 'vocab.data'), 'rb') as fin:
vocab = pickle.load(fin)
assert len(args.test_files) > 0, '找不到测试文件.'
brc_data = BRCDataset(args.max_p_num,
args.max_p_len,
args.max_q_len,
test_files=args.test_files)
logger.info('把文本转化为id序列...')
brc_data.convert_to_ids(vocab)
logger.info('重载模型...')
rc_model = RCModel(vocab, args)
rc_model.restore(model_dir=args.model_dir, model_prefix=args.algo)
logger.info('预测测试集的答案...')
test_batches = brc_data.gen_mini_batches('test',
args.batch_size,
pad_id=vocab.get_id(
vocab.pad_token),
shuffle=False)
rc_model.evaluate(test_batches,
result_dir=args.result_dir,
result_prefix='test.predicted')
def predict(args):
"""
predicts answers for test files
"""
logger = logging.getLogger("Military AI")
logger.info('Load data_set and vocab...')
mai_data = MilitaryAiDataset(args.train_files,
args.train_raw_files,
args.test_files,
args.test_raw_files,
args.char_embed_file,
args.token_embed_file,
args.elmo_dict_file,
args.elmo_embed_file,
char_min_cnt=1,
token_min_cnt=3)
logger.info('Assigning embeddings...')
if not args.use_embe:
mai_data.token_vocab.randomly_init_embeddings(args.embed_size)
mai_data.char_vocab.randomly_init_embeddings(args.embed_size)
logger.info('Restoring the model...')
rc_model = RCModel(mai_data.char_vocab, mai_data.token_vocab,
mai_data.flag_vocab, mai_data.elmo_vocab, args)
rc_model.restore(model_dir=args.model_dir,
model_prefix=args.algo + args.suffix)
logger.info('Predicting answers for test set...')
test_batches = mai_data.gen_mini_batches('test',
args.batch_size,
shuffle=False)
rc_model.evaluate(test_batches,
result_dir=args.result_dir,
result_prefix='test.predicted')
def train(args):
"""
trains the reading comprehension model
"""
logger = logging.getLogger("brc")
logger.info('Loading vocab...')
with open(os.path.join(args.vocab_dir, 'vocab.pkl'), 'rb') as fin:
vocab = pickle.load(fin)
fin.close()
pad_id = vocab.get_id(vocab.pad_token)
brc_data = BRCDataset(args.max_p_num, args.max_p_len, args.max_q_len,
args.prepared_dir, args.train_files, args.dev_files,
args.test_files)
logger.info('Converting text into ids...')
brc_data.convert_to_ids(vocab)
g = tf.Graph()
with g.as_default():
rc_model = RCModel(vocab.embeddings, pad_id, args)
del vocab
# Train
with tf.name_scope("Train"):
logger.info('Training the model...')
rc_model.train(brc_data,
args.epochs,
args.batch_size,
save_dir=args.result_dir,
save_prefix='test.predicted',
dropout_keep_prob=args.dropout_keep_prob)
tf.summary.FileWriter(args.summary_dir, g).close()
with tf.name_scope('Valid'):
assert len(args.dev_files) > 0, 'No dev files are provided.'
logger.info('Evaluating the model on dev set...')
dev_batches = brc_data.gen_mini_batches('dev',
args.batch_size,
pad_id=pad_id,
shuffle=False)
dev_loss, dev_bleu_rouge = rc_model.evaluate(
dev_batches,
result_dir=args.result_dir,
result_prefix='dev.predicted')
logger.info('Loss on dev set: {}'.format(dev_loss))
logger.info('Result on dev set: {}'.format(dev_bleu_rouge))
logger.info('Predicted answers are saved to {}'.format(
os.path.join(args.result_dir)))
with tf.name_scope('Test'):
assert len(args.test_files) > 0, 'No test files are provided.'
logger.info('Predicting answers for test set...')
test_batches = brc_data.gen_mini_batches('test',
args.batch_size,
pad_id=pad_id,
shuffle=False)
rc_model.evaluate(test_batches,
result_dir=args.result_dir,
result_prefix='test.predicted')
def evaluate(args):
"""
evaluate the trained model on dev files
"""
logger = logging.getLogger("brc")
logger.info('Load data_set and vocab...')
# with open(os.path.join(args.vocab_dir, 'vocab.data'), 'rb') as fin:
with open(args.vocab_path, 'rb') as fin:
vocab = pickle.load(fin)
assert len(args.dev_files) > 0, 'No dev files are provided.'
brc_data = BRCDataset(args.max_p_num,
args.max_p_len,
args.max_q_len,
dev_files=args.dev_files)
logger.info('Converting text into ids...')
brc_data.convert_to_ids(vocab)
logger.info('Restoring the model...')
rc_model = RCModel(vocab, args)
rc_model.restore(model_dir=args.model_dir, model_prefix=args.algo)
logger.info('Evaluating the model on dev set...')
dev_batches = brc_data.gen_mini_batches('dev',
args.batch_size,
pad_id=vocab.get_id(
vocab.pad_token),
shuffle=False)
dev_loss, dev_bleu_rouge = rc_model.evaluate(dev_batches,
result_dir=args.result_dir,
result_prefix='dev.predicted')
logger.info('Loss on dev set: {}'.format(dev_loss))
logger.info('Result on dev set: {}'.format(dev_bleu_rouge))
logger.info('Predicted answers are saved to {}'.format(
os.path.join(args.result_dir)))
def evaluate(args):
"""
对训练好的模型进行验证
"""
logger = logging.getLogger("brc")
logger.info('加wudi...')
logger.info('加载数据集和词汇表...')
with open(os.path.join(args.vocab_dir, 'vocab.data'), 'rb') as fin:
vocab = pickle.load(fin)
assert len(args.dev_files) > 0, '找不到验证文件.'
brc_data = BRCDataset(args.max_p_num,
args.max_p_len,
args.max_q_len,
dev_files=args.dev_files)
logger.info('把文本转化为id序列...')
brc_data.convert_to_ids(vocab)
logger.info('重载模型...')
rc_model = RCModel(vocab, args)
rc_model.restore(model_dir=args.model_dir, model_prefix=args.algo)
logger.info('验证模型...')
dev_batches = brc_data.gen_mini_batches('dev',
args.batch_size,
pad_id=vocab.get_id(
vocab.pad_token),
shuffle=False)
dev_loss, dev_bleu_rouge = rc_model.evaluate(dev_batches,
result_dir=args.result_dir,
result_prefix='dev.predicted')
logger.info('验证集上的损失为: {}'.format(dev_loss))
logger.info('验证集的结果: {}'.format(dev_bleu_rouge))
logger.info('预测的答案证保存到 {}'.format(os.path.join(args.result_dir)))
def predict(args):
"""
predicts answers for test files
"""
logger = logging.getLogger("brc")
logger.info('Load data_set and vocab...')
with open(os.path.join(args.vocab_dir, 'vocab.data'), 'rb') as fin:
vocab = pickle.load(fin)
assert len(args.test_files) > 0, 'No test files are provided.'
brc_data = BRCDataset(args.max_p_num, args.max_p_len, args.max_q_len,
test_files=args.test_files)
logger.info('Converting text into ids...')
brc_data.convert_to_ids(vocab)
logger.info('Restoring the model...')
rc_model = RCModel(vocab, args)
rc_model.restore(model_dir=args.model_dir, model_prefix=args.algo)
logger.info('Predicting answers for test set...')
test_batches = brc_data.gen_mini_batches('test', args.batch_size,
pad_id=vocab.get_id(vocab.pad_token), shuffle=False)
rc_model.evaluate(test_batches,
result_dir=args.result_dir, result_prefix='test.predicted')
def predict(args):
"""
predicts answers for test files
"""
logger = logging.getLogger("brc")
logger.info('Load data_set and vocab...')
with open(os.path.join(args.vocab_dir, 'vocab.data'), 'rb') as fin:
vocab = pickle.load(fin)
assert len(args.test_files) > 0, 'No test files are provided.'
brc_data = BRCDataset(args.max_p_num,
args.max_p_len,
args.max_q_len,
args.max_word_len,
test_files=args.test_files)
logger.info('Converting text into ids...')
brc_data.convert_to_ids(vocab)
logger.info('Restoring the model...')
rc_model = RCModel(vocab, args)
rc_model.restore(model_dir=args.model_dir, model_prefix=args.algo)
rc_model.finalize()
# 增加完所有操作后采用sess.graph.finalize()
# 来使得整个graph变为只读的
# 注意：tf.train.Saver()
# 也算是往graph中添加node， 所以也必须放在finilize前
# 但是，，tf.train.Saver()
# 只会存储
# 在该Saver声明时已经存在的变量！！！
logger.info('Predicting answers for test set...')
test_batches = brc_data.gen_mini_batches('test',
args.batch_size,
pad_id=vocab.get_id(
vocab.pad_token),
shuffle=False)
rc_model.evaluate(test_batches,
result_dir=args.result_dir,
result_prefix='test.predicted')
def evaluate(args):
"""
evaluate the trained model on dev files
"""
logger = logging.getLogger("brc")
logger.info('Load data_set and vocab...')
with open(os.path.join(args.vocab_dir, 'vocab.data'), 'rb') as fin:
vocab = pickle.load(fin)
assert len(args.dev_files) > 0, 'No dev files are provided.'
brc_data = BRCDataset(args.max_p_num, args.max_p_len, args.max_q_len, dev_files=args.dev_files)
logger.info('Converting text into ids...')
brc_data.convert_to_ids(vocab)
logger.info('Restoring the model...')
rc_model = RCModel(vocab, args)
rc_model.restore(model_dir=args.model_dir, model_prefix=args.algo)
logger.info('Evaluating the model on dev set...')
dev_batches = brc_data.gen_mini_batches('dev', args.batch_size,
pad_id=vocab.get_id(vocab.pad_token), shuffle=False)
dev_loss, dev_bleu_rouge = rc_model.evaluate(
dev_batches, result_dir=args.result_dir, result_prefix='dev.predicted')
logger.info('Loss on dev set: {}'.format(dev_loss))
logger.info('Result on dev set: {}'.format(dev_bleu_rouge))
logger.info('Predicted answers are saved to {}'.format(os.path.join(args.result_dir)))
def evaluate(args):
"""
evaluate the trained model on dev files
"""
logger = logging.getLogger("Military AI")
logger.info('Load data_set and vocab...')
mai_data = MilitaryAiDataset(args.train_files,
args.train_raw_files,
args.test_files,
args.test_raw_files,
args.char_embed_file,
args.token_embed_file,
args.elmo_dict_file,
args.elmo_embed_file,
char_min_cnt=1,
token_min_cnt=3)
logger.info('Assigning embeddings...')
if not args.use_embe:
mai_data.token_vocab.randomly_init_embeddings(args.embed_size)
mai_data.char_vocab.randomly_init_embeddings(args.embed_size)
logger.info('Restoring the model...')
rc_model = RCModel(mai_data.char_vocab, mai_data.token_vocab,
mai_data.flag_vocab, mai_data.elmo_vocab, args)
rc_model.restore(model_dir=args.model_dir,
model_prefix=args.algo + args.suffix)
logger.info('Evaluating the model on dev set...')
dev_batches = mai_data.gen_mini_batches('dev',
args.batch_size,
shuffle=False)
dev_loss, dev_main_loss, dev_bleu_rouge = rc_model.evaluate(
dev_batches, result_dir=args.result_dir, result_prefix='dev.predicted')
logger.info('Loss on dev set: {}'.format(dev_main_loss))
logger.info('Result on dev set: {}'.format(dev_bleu_rouge))
logger.info('Predicted answers are saved to {}'.format(
os.path.join(args.result_dir)))
def predict(args):
"""
predicts answers for test files
"""
logger = logging.getLogger("brc")
logger.info('Load data_set and vocab...')
with open(args.vocab_path, 'rb') as fin:
vocab = pickle.load(fin)
assert len(args.test_files) > 0, 'No test files are provided.'
brc_data = BRCDataset(args.algo,
args.max_p_num,
args.max_p_len,
args.max_q_len,
args.max_a_len,
test_files=args.test_files)
logger.info('Converting text into ids...')
brc_data.convert_to_ids(vocab)
logger.info('Restoring the model...')
rc_model = RCModel(vocab, args)
rc_model.restore(model_dir=args.model_dir, model_prefix=args.algo)
logger.info('Predicting answers for test set...')
test_batches = brc_data.gen_mini_batches('test',
args.batch_size,
pad_id=vocab.get_id(
vocab.pad_token),
shuffle=False)
if args.algo == 'YESNO':
qa_resultPath = args.test_files[0] #只会有一个文件！
(filepath, tempfilename) = os.path.split(qa_resultPath)
(qarst_filename, extension) = os.path.splitext(tempfilename)
result_prefix = qarst_filename
else:
result_prefix = 'test.predicted.qa'
rc_model.evaluate(test_batches,
result_dir=args.result_dir,
result_prefix=result_prefix)
if args.algo == 'YESNO': #将YESNO结果合并入QA结果
qa_resultPath = args.test_files[0] #只会有一个文件！
yesno_resultPath = args.result_dir + '/' + result_prefix + '.YESNO.json'
out_file_path = args.result_dir + '/' + result_prefix + '.134.class.' + str(
args.run_id) + '.json'
#首先载入YESNO部分的预测结果
yesno_records = {}
with open(yesno_resultPath, 'r') as f_in:
for line in f_in:
sample = json.loads(line)
yesno_records[sample['question_id']] = line
total_rst_num = 0
with open(qa_resultPath, 'r') as f_in:
with open(out_file_path, 'w') as f_out:
for line in f_in:
total_rst_num += 1
sample = json.loads(line)
if sample['question_id'] in yesno_records:
line = yesno_records[sample['question_id']]
f_out.write(line)
print('total rst num : ', total_rst_num)
print('yes no label combining done!')