Created by: MsFive5

Got this error when making an inference using paddle

WARNING: Logging before InitGoogleLogging() is written to STDERR
F0814 16:02:02.693090 32322 ClassRegistrar.h:65] Check failed: mapGet(type, creatorMap_, &creator) Unknown class type: data
*** Check failure stack trace: ***
Aborted (core dumped)

Here is how I write reader and inference

def reader(query_seg, word_dict):
    UNK_ID = word_dict["<UNK>"]
    for line in f:
        word_ids = [
            word_dict.get(w, UNK_ID)
            for w in query_seg.split()
        ]
        yield word_ids

def infer(query_seg, model_path, label_dict_path, word_dict_path):
    word_dict = load_dict(word_dict_path)
    label_reverse_dict = load_reverse_dict(label_dict_path)
    dict_dim = len(word_dict)
    class_num = len(label_reverse_dict)
    topology = stacked_lstm_net
    prob_layer = topology(dict_dim, class_num, is_infer=True)
    parameters = paddle.parameters.Parameters.from_tar(
        gzip.open(model_path, "r"))
   inferer = paddle.inference.Inference(
        output_layer=prob_layer, parameters=parameters)
    test_vector = reader(query_seg, word_dict)
    probs = inferer.infer(input=test_vector, field=["value"])
    label = label_reverse_dict[prob.argmax()]
    return label

Found out that core dumped happens here

 inferer = paddle.inference.Inference(
        output_layer=prob_layer, parameters=parameters)

I write this inferer the same as the demo. Please help!