How to fix the ValueError

[Myprojectjoy]

Environment details

If you are already running DeepEcho, please indicate the following details about the environment in
which you are running it:

• DeepEcho version: deepecho 0.3.0.post1
• Python version: 3.8
• Operating System: Google colab, Jupyter Notebook

Problem description

What I already tried

In the time series dataset, there is no entity_column and context_column.

Sequesnce_indext is the date column, and I have chosen "segment_size = 1 "

I have a data column with a float value, including a negative value, replaced with a 0.00 value.
There is no NaN value in the entire data frame.

df.isnull().values.any()
Output: False

In the data_tyeps section, I have selected that column as,

1. data_types = {'hour': 'count',
   'minute': 'categorical',
   'month' : 'count',
   'year' : 'count',
   'dayofmonth' : 'categorical',
   'Power' : 'count',
   }

2. model = PARModel(epochs=10, cuda=True)

3. Learn a model from the data

model.fit(
data=df,
segment_size = 1,
data_types=data_types,
sequence_index='Date'
)

Paste the command(s) you ran and the output.
If there was a crash, please include the traceback here.

ValueError: Expected parameter loc (Tensor of shape (2,)) of distribution Normal(loc: torch.Size([2]), scale: torch.Size([2])) to satisfy the constraint Real(), but found invalid values:
tensor([nan, nan], device='cuda:0', grad_fn=)