Created by: bbingy

Load pretraining parameters from /home/lby/ERNIE_stable/params. /paddle/paddle/fluid/operators/math/cross_entropy.cu:40 Assertion label[i] >= 0 && label[i] < D || label[i] == ignore_index failed. Traceback (most recent call last): File "run_classifier.py", line 276, in main(args) File "run_classifier.py", line 195, in main train_exe.run(fetch_list=[]) File "/home/lby/.conda/envs/paddlepaddle/lib/python3.6/site-packages/paddle/fluid/parallel_executor.py", line 303, in run

1 0x7f9b22cf2759p paddle::platform::EnforceNotMet::EnforceNotMet(std::__exception_ptr::exception_ptr, char const*, int) + 137 2 0x7f9b247840c6p 3 0x7f9b247a67d4p paddle::platform::TemporaryAllocator::Release(std::function<void ()> const&) + 100 4 0x7f9b247860e1p paddle::platform::CUDADeviceContext::Wait() const + 113 5 0x7f9b24506a68p paddle::framework::details::ScopeBufferedSSAGraphExecutor::WaitComputationalStreams() + 120 6 0x7f9b24506072p paddle::framework::details::ScopeBufferedSSAGraphExecutor::Run(std::vector<std::string, std::allocatorstd::string > const&) + 1794 7 0x7f9b22e2d872p paddle::framework::ParallelExecutor::Run(std::vector<std::string, std::allocatorstd::string > const&, std::string const&) + 562 8 0x7f9b22ce2c7ep 9 0x7f9b22d1d79ep 10 0x55838e22b744p _PyCFunction_FastCallDict + 340 11 0x55838e2b257ep 12 0x55838e2d738ap _PyEval_EvalFrameDefault + 778 13 0x55838e2abbfep 14 0x55838e2ac771p 15 0x55838e2b2505p 16 0x55838e2d8147p _PyEval_EvalFrameDefault + 4295 17 0x55838e2ac53bp 18 0x55838e2b2505p 19 0x55838e2d738ap _PyEval_EvalFrameDefault + 778 20 0x55838e2ad289p PyEval_EvalCodeEx + 809 21 0x55838e2ae01cp PyEval_EvalCode + 28 22 0x55838e3303c4p 23 0x55838e3307c1p PyRun_FileExFlags + 161 24 0x55838e3309c3p PyRun_SimpleFileExFlags + 451 25 0x55838e3344b3p Py_Main + 1555 26 0x55838e1fd02ep main + 238 27 0x7f9b5c84db97p __libc_start_main + 231 28 0x55838e2dde0ep

多分类会报这样的错误，请问是否可以使用预训练好的模型进行多分类呢？可以的话应该修改哪里呢？