Created by: linrongyi

使用预测工具预测sparse_binary_vector的instance, 执行network.forwardTest(in_arg) 出错

代码

# -*- coding: utf-8 -*-

import sys

import numpy as np

from py_paddle import swig_paddle
from py_paddle import DataProviderConverter

from paddle.trainer.PyDataProvider2 import *

#from paddle.trainer.PyDataProviderWrapper import DenseSlot, IndexSlot
from paddle.trainer.config_parser import parse_config

swig_paddle.initPaddle("--use_gpu=0")

def get_slots():
    slots = [sparse_binary_vector(800000)]
    for i in xrange(8):
        slots.append(sparse_binary_vector(800000))
        slots.append(sparse_binary_vector(800000))
    return slots

slots = get_slots()

conf = parse_config("./output/pass-00000/predictor_config.conf", "")
network = swig_paddle.GradientMachine.createFromConfigProto(conf.model_config)
converter = DataProviderConverter(slots)

instance_ = [[1]] * 17
in_arg = converter([instance_])
network.forwardTest(in_arg)

错误提示

I1105 20:46:14.398023 11953 Util.cpp:155] commandline:  --use_gpu=0 
I1105 20:46:14.398080 11953 Util.cpp:130] Calling runInitFunctions
I1105 20:46:14.398355 11953 Util.cpp:143] Call runInitFunctions done.
F1105 20:46:14.660048 11953 SparseRowMatrix.h:63] Check failed: globalIndices_[row] != kUnusedId_ (4294967295 vs. 4294967295) 
*** Check failure stack trace: ***
    @     0x7f45a5946a3d  google::LogMessage::Fail()
    @     0x7f45a594aed7  google::LogMessage::SendToLog()
    @     0x7f45a5948d39  google::LogMessage::Flush()
    @     0x7f45a594903d  google::LogMessageFatal::~LogMessageFatal()
    @     0x7f45a725eea2  paddle::CpuMatrix::mul<>()
    @     0x7f45a7256237  paddle::CpuMatrix::mul()
    @     0x7f45a70e716e  paddle::FullyConnectedLayer::forward()
    @     0x7f45a70aad64  paddle::NeuralNetwork::forward()
    @     0x7f45a701d955  _wrap_GradientMachine_forward
    @     0x7f45b2e273a3  PyEval_EvalFrameEx
    @     0x7f45b2e29130  PyEval_EvalCodeEx
    @     0x7f45b2e274a1  PyEval_EvalFrameEx
    @     0x7f45b2e29130  PyEval_EvalCodeEx
    @     0x7f45b2e274a1  PyEval_EvalFrameEx
    @     0x7f45b2e29130  PyEval_EvalCodeEx
    @     0x7f45b2e29242  PyEval_EvalCode
    @     0x7f45b2e4362c  run_mod
    @     0x7f45b2e43700  PyRun_FileExFlags
    @     0x7f45b2e44c0c  PyRun_SimpleFileExFlags
    @     0x7f45b2e564cc  Py_Main
    @       0x318ae1ecdd  (unknown)
Thread [139937329645312] Forwarding query_vec, site7, title7, site6, title6, site5, title5, site4, title4, site3, title3, site2, title2, site1, title1, site0, title0, query, 
*** Aborted at 1478349974 (unix time) try "date -d @1478349974" if you are using GNU date ***
PC: @       0x318ae328a5 (unknown)
*** SIGABRT (@0x1f500002eb1) received by PID 11953 (TID 0x7f45b2d30700) from PID 11953; stack trace: ***
    @       0x318b20f500 (unknown)
    @       0x318ae328a5 (unknown)
    @       0x318ae34085 (unknown)
    @     0x7f45a5953246 google::FindSymbol()

版本号, 93e4d0cc.

麻烦 @reyoung 看一下?