CUDA error: device-side assert triggered

[Raviteja-banda]

I have been trying to run the speaker_id.py file(for a customised librispeech dataset) but i was stuck at this error from the past few days. I have even checked if there's any mismatch between the number of labels and the output neurons but they are fine. I've changed the config file as below:
class_lay=251
batch_size=8
N_batches=251 (Since total number of train files(excluding validation data) = 2008)

[windowing]
fs=16000
cw_len=200
cw_shift=10

[cnn]
cnn_N_filt=80,60,60
cnn_len_filt=251,5,5
cnn_max_pool_len=3,3,3
cnn_use_laynorm_inp=True
cnn_use_batchnorm_inp=False
cnn_use_laynorm=True,True,True
cnn_use_batchnorm=False,False,False
cnn_act=leaky_relu,leaky_relu,leaky_relu
cnn_drop=0.0,0.0,0.0

[dnn]
fc_lay=2048,2048,2048
fc_drop=0.0,0.0,0.0
fc_use_laynorm_inp=True
fc_use_batchnorm_inp=False
fc_use_batchnorm=True,True,True
fc_use_laynorm=False,False,False
fc_act=leaky_relu,leaky_relu,leaky_relu

[class]
class_lay=251
class_drop=0.0
class_use_laynorm_inp=False
class_use_batchnorm_inp=False
class_use_batchnorm=False
class_use_laynorm=False
class_act=softmax

I have also changed the data paths in the config file.

Below is my stack trace of the error:

C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Loss.cu:242: block: [0,0,0], thread: [0,0,0] Assertion t >= 0 && t < n_classes failed.
C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Loss.cu:242: block: [0,0,0], thread: [1,0,0] Assertion t >= 0 && t < n_classes failed.
C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Loss.cu:242: block: [0,0,0], thread: [2,0,0] Assertion t >= 0 && t < n_classes failed.
C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Loss.cu:242: block: [0,0,0], thread: [3,0,0] Assertion t >= 0 && t < n_classes failed.
C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Loss.cu:242: block: [0,0,0], thread: [4,0,0] Assertion t >= 0 && t < n_classes failed.
C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Loss.cu:242: block: [0,0,0], thread: [5,0,0] Assertion t >= 0 && t < n_classes failed.
C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Loss.cu:242: block: [0,0,0], thread: [6,0,0] Assertion t >= 0 && t < n_classes failed.
C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Loss.cu:242: block: [0,0,0], thread: [7,0,0] Assertion t >= 0 && t < n_classes failed.
Traceback (most recent call last):
File "D:\SincNet-librispeech\speaker_id.py", line 245, in
loss.backward()
File "C:\ProgramData\Anaconda3\lib\site-packages\torch_tensor.py", line 487, in backward
torch.autograd.backward(
File "C:\ProgramData\Anaconda3\lib\site-packages\torch\autograd_init_.py", line 197, in backward
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
RuntimeError: CUDA error: device-side assert triggered

Any possible solutions?