# -*- coding: utf-8 -*-
"""
@author: mwahdan
"""
from dialognlu import TransformerNLU
from dialognlu.readers.goo_format_reader import Reader
# model_path = "../saved_models/joint_distilbert_model"
model_path = "../saved_models/joint_trans_xlnet_model"
print("Loading model ...")
nlu = TransformerNLU.load(model_path)
print("Loading dataset ...")
test_path = "../data/snips/test"
test_dataset = Reader.read(test_path)
print("Evaluating model ...")
token_f1_score, tag_f1_score, report, acc = nlu.evaluate(test_dataset)
print('Slot Classification Report:', report)
print('Slot token f1_score = %f' % token_f1_score)
print('Slot tag f1_score = %f' % tag_f1_score)
print('Intent accuracy = %f' % acc)
# -*- coding: utf-8 -*-
"""
@author: mwahdan
"""
from dialognlu import BertNLU
from dialognlu.readers.goo_format_reader import Reader
train_path = "../data/snips/train"
val_path = "../data/snips/valid"
train_dataset = Reader.read(train_path)
val_dataset = Reader.read(val_path)
save_path = "../saved_models/joint_bert_model"
epochs = 1 #3
batch_size = 64
config = {"model_type": "bert"}
nlu = BertNLU.from_config(config)
nlu.train(train_dataset, val_dataset, epochs, batch_size)
print("Saving ...")
nlu.save(save_path)
print("Done")
args = parser.parse_args()
train_data_folder_path = args.train
val_data_folder_path = args.val
save_folder_path = args.save
epochs = args.epochs
batch_size = args.batch
start_model_folder_path = args.model
pretrained_model_name_or_path = args.trans
from_pt = args.from_pt
cache_dir = args.cache_dir
if start_model_folder_path is None and pretrained_model_name_or_path is None:
raise argparse.ArgumentTypeError(
"Either --model OR --trans should be provided")
print('Reading data ...')
train_dataset = Reader.read(train_data_folder_path)
val_dataset = Reader.read(val_data_folder_path)
if start_model_folder_path is None:
config = {
"cache_dir": cache_dir,
"pretrained_model_name_or_path": pretrained_model_name_or_path,
"from_pt": from_pt,
"num_bert_fine_tune_layers": 10,
"intent_loss_weight": 1.0,
"slots_loss_weight": 3.0,
}
nlu = TransformerNLU.from_config(config)
else:
nlu = TransformerNLU.load(start_model_folder_path)
required=True)
parser.add_argument('--data',
'-d',
help='Path to data in Goo et al format',
type=str,
required=True)
parser.add_argument('--batch',
'-bs',
help='Batch size',
type=int,
default=128,
required=False)
args = parser.parse_args()
model_path = args.model
data_folder_path = args.data
batch_size = args.batch
print("Loading model ...")
nlu = AutoNLU.load(model_path)
print("Loading dataset ...")
test_dataset = Reader.read(data_folder_path)
print("Evaluating model ...")
token_f1_score, tag_f1_score, report, acc = nlu.evaluate(test_dataset)
print('Slot Classification Report:', report)
print('Slot token f1_score = %f' % token_f1_score)
print('Slot tag f1_score = %f' % tag_f1_score)
print('Intent accuracy = %f' % acc)