def test_NMT_Unidir_GRU_ConditionalGRU():
params = load_tests_params()
# Current test params: Single layered GRU - ConditionalGRU
params['BIDIRECTIONAL_ENCODER'] = False
params['N_LAYERS_ENCODER'] = 1
params['BIDIRECTIONAL_DEEP_ENCODER'] = False
params['ENCODER_RNN_TYPE'] = 'GRU'
params['DECODER_RNN_TYPE'] = 'ConditionalGRU'
params['N_LAYERS_DECODER'] = 1
params['REBUILD_DATASET'] = True
dataset = build_dataset(params)
params['INPUT_VOCABULARY_SIZE'] = dataset.vocabulary_len[params['INPUTS_IDS_DATASET'][0]]
params['OUTPUT_VOCABULARY_SIZE'] = dataset.vocabulary_len[params['OUTPUTS_IDS_DATASET'][0]]
params['MODEL_NAME'] = \
params['TASK_NAME'] + '_' + params['SRC_LAN'] + params['TRG_LAN'] + '_' + params['MODEL_TYPE'] + \
'_src_emb_' + str(params['SOURCE_TEXT_EMBEDDING_SIZE']) + \
'_bidir_' + str(params['BIDIRECTIONAL_ENCODER']) + \
'_enc_' + params['ENCODER_RNN_TYPE'] + '_*' + str(params['N_LAYERS_ENCODER']) + '_' + str(
params['ENCODER_HIDDEN_SIZE']) + \
'_dec_' + params['DECODER_RNN_TYPE'] + '_*' + str(params['N_LAYERS_DECODER']) + '_' + str(
params['DECODER_HIDDEN_SIZE']) + \
'_deepout_' + '_'.join([layer[0] for layer in params['DEEP_OUTPUT_LAYERS']]) + \
'_trg_emb_' + str(params['TARGET_TEXT_EMBEDDING_SIZE']) + \
'_' + params['OPTIMIZER'] + '_' + str(params['LR'])
params['STORE_PATH'] = K.backend() + '_test_train_models/' + params['MODEL_NAME'] + '/'
# Test several NMT-Keras utilities: train, sample, sample_ensemble, score_corpus...
print ("Training model")
train_model(params)
params['RELOAD'] = 1
print ("Done")
print ("Applying model")
apply_NMT_model(params)
print ("Done")
parser = argparse.ArgumentParser('Parser for unit testing')
parser.dataset = params['DATASET_STORE_PATH'] + '/Dataset_' + params['DATASET_NAME'] + '_' + params['SRC_LAN'] + params['TRG_LAN'] + '.pkl'
parser.text = params['DATA_ROOT_PATH'] + '/' + params['TEXT_FILES']['val'] + params['SRC_LAN']
parser.splits = ['val']
parser.config = params['STORE_PATH'] + '/config.pkl'
parser.models = [params['STORE_PATH'] + '/epoch_' + str(2)]
parser.verbose = 0
parser.dest = None
parser.source = params['DATA_ROOT_PATH'] + '/' + params['TEXT_FILES']['val'] + params['SRC_LAN']
parser.target = params['DATA_ROOT_PATH'] + '/' + params['TEXT_FILES']['val'] + params['TRG_LAN']
parser.weights = []
for n_best in [True, False]:
parser.n_best = n_best
print ("Sampling with n_best = %s "% str(n_best))
sample_ensemble(parser, params)
print ("Done")
print ("Scoring corpus")
score_corpus(parser, params)
print ("Done")
def test_transformer():
params = load_tests_params()
# Current test params: Transformer
params['N_LAYERS_ENCODER'] = 2
params['N_LAYERS_DECODER'] = 2
params['MULTIHEAD_ATTENTION_ACTIVATION'] = 'relu'
params['MODEL_SIZE'] = 32
params['FF_SIZE'] = MODEL_SIZE * 4
params['N_HEADS'] = 2
params['REBUILD_DATASET'] = True
dataset = build_dataset(params)
params['INPUT_VOCABULARY_SIZE'] = dataset.vocabulary_len[
params['INPUTS_IDS_DATASET'][0]]
params['OUTPUT_VOCABULARY_SIZE'] = dataset.vocabulary_len[
params['OUTPUTS_IDS_DATASET'][0]]
params['MODEL_NAME'] =\
params['TASK_NAME'] + '_' + params['SRC_LAN'] + params['TRG_LAN'] + '_' + params['MODEL_TYPE'] + \
'_model_size_' + str(params['MODEL_SIZE']) + \
'_ff_size_' + str(params['FF_SIZE']) + \
'_num_heads_' + str(params['N_HEADS']) + \
'_encoder_blocks_' + str(params['N_LAYERS_ENCODER']) + \
'_decoder_blocks_' + str(params['N_LAYERS_DECODER']) + \
'_deepout_' + '_'.join([layer[0] for layer in params['DEEP_OUTPUT_LAYERS']]) + \
'_' + params['OPTIMIZER'] + '_' + str(params['LR'])
params['STORE_PATH'] = K.backend(
) + '_test_train_models/' + params['MODEL_NAME'] + '/'
# Test several NMT-Keras utilities: train, sample, sample_ensemble, score_corpus...
train_model(params)
params['RELOAD'] = 2
apply_NMT_model(params)
parser = argparse.ArgumentParser('Parser for unit testing')
parser.dataset = params['DATASET_STORE_PATH'] + '/Dataset_' + params[
'DATASET_NAME'] + '_' + params['SRC_LAN'] + params['TRG_LAN'] + '.pkl'
parser.text = params['DATA_ROOT_PATH'] + '/' + params['TEXT_FILES'][
'val'] + params['SRC_LAN']
parser.splits = ['val']
parser.config = params['STORE_PATH'] + '/config.pkl'
parser.models = [params['STORE_PATH'] + '/epoch_' + str(2)]
parser.verbose = 0
parser.dest = None
parser.source = params['DATA_ROOT_PATH'] + '/' + params['TEXT_FILES'][
'val'] + params['SRC_LAN']
parser.target = params['DATA_ROOT_PATH'] + '/' + params['TEXT_FILES'][
'val'] + params['TRG_LAN']
parser.weights = []
for n_best in [True, False]:
parser.n_best = n_best
sample_ensemble(parser, params)
score_corpus(parser, params)
