with open(args.model_filename + ".params", mode="r") as in_file:
params = json.load(in_file)
print("-- Loading index")
with open(args.model_filename + ".index", mode="rb") as in_file:
index = pickle.load(in_file)
token2id = index["token2id"]
id2token = index["id2token"]
label2id = index["label2id"]
id2label = index["id2label"]
num_tokens = len(token2id)
num_labels = len(label2id)
print("-- Loading test set")
print("-- Loading test set")
test_labels, test_padded_sentences, test_img_names, test_original_sentences = load_ic_dataset(
args.test_filename, token2id, label2id)
print("-- Loading images")
image_reader = ImageReader(args.img_names_filename,
args.img_features_filename)
print("-- Restoring model")
sentence_input = tf.placeholder(tf.int32, (None, None),
name="sentence_input")
img_features_input = tf.placeholder(
tf.float32,
(None, params["num_img_features"], params["img_features_size"]),
name="img_features_input")
label_input = tf.placeholder(tf.int32, (None, ), name="label_input")
dropout_input = tf.placeholder(tf.float32, name="dropout_input")
logits = build_bottom_up_top_down_ic_model(
print("Params saved to: {}".format(args.model_save_filename +
".params"))
with open(args.model_save_filename + ".index", mode="wb") as out_file:
pickle.dump(
{
"token2id": token2id,
"id2token": id2token,
"label2id": label2id,
"id2label": id2label
}, out_file)
print("Index saved to: {}".format(args.model_save_filename +
".index"))
print("-- Loading training set")
train_labels, train_sentences, _, _ = load_ic_dataset(
args.train_filename, token2id, label2id)
print("-- Loading development set")
dev_labels, dev_sentences, _, _ = load_ic_dataset(args.dev_filename,
token2id, label2id)
print("-- Building model")
sentence_input = tf.placeholder(tf.int32, (None, None),
name="sentence_input")
label_input = tf.placeholder(tf.int32, (None, ), name="label_input")
dropout_input = tf.placeholder(tf.float32, name="dropout_input")
logits = build_simple_blind_ic_model(sentence_input, dropout_input,
num_tokens, num_labels, embeddings,
args.embeddings_size,
args.train_embeddings,
args.rnn_hidden_size,
with open(args.model_save_filename + ".index", mode="wb") as out_file:
pickle.dump(
{
"token2id": token2id,
"id2token": id2token,
"vte_label2id": vte_label2id,
"vte_id2label": vte_id2label,
"ic_label2id": ic_label2id,
"ic_id2label": ic_id2label
}, out_file)
print("Index saved to: {}".format(args.model_save_filename +
".index"))
print("-- Loading training set")
ic_train_labels, ic_train_sentences, ic_train_img_names, _ = load_ic_dataset(
args.ic_train_filename, token2id, ic_label2id)
print("-- Loading training set")
vte_train_labels, vte_train_premises, vte_train_hypotheses, vte_train_img_names, _, _ =\
load_vte_dataset(
args.vte_train_filename,
token2id,
vte_label2id
)
print("-- Loading development set")
vte_dev_labels, vte_dev_premises, vte_dev_hypotheses, vte_dev_img_names, _, _ =\
load_vte_dataset(
args.vte_dev_filename,
token2id,
vte_label2id