01. PyTorch Workflow Fundamentals Timestand - 5:34:45sec

[Gauravjake]

here are the code and I got the error that object has no attribute 'weights'

from torch import nn

Create lines repression model class

class LinearRegressionModel(nn.Module): # -> almost everything in Pytorch inherhits from nn.Module
def init(self):
super().init()
self.weights = nn.Parameter(torch.randn(1, # <- start with a random weight and try to adjust it to the ideal weight
requires_grad=True, # <- can this parameter be updated via gradient descent?
dtype=torch.float)) # <- Pytorch loves the datatype torch float32

self.bias = nn.Parameter(torch.randn(1, #<- start with a random bias and try to adjust it to the ideal bias
require_grad=True, # <- can this parameter be update via gradient descent?
dtype=torch.float)) # <- Python loves the datatype torch.float32

Forward method to define the computation is the node:

def forward(self, x: torch.Tensor) -> torch.Tensor: # <- "x" is the input data
return self.weights * x + self.bias # this is the linear regression formula
``

AttributeError Traceback (most recent call last)
in
2 with torch.inference_mode():
3
----> 4 y_preds = model_0(x_test)
5

2 frames
/usr/local/lib/python3.8/dist-packages/torch/nn/modules/module.py in getattr(self, name)
1263 if name in modules:
1264 return modules[name]
-> 1265 raise AttributeError("'{}' object has no attribute '{}'".format(
1266 type(self).name, name))
1267

AttributeError: 'LinearRegressionModel' object has no attribute 'weights'

.

**** Can any please explain me what is going on here. I search this in stack overflow but still can fix my problem and doubt.

[BoolinBood]

on line 3, try changing super().init() to super().init(), I don't know this might fix the problem but let give it a try.

[PatTheAtak]

Hey @Gauravjake , I hope you're well.
So, yeah, I've gone through the piece of code you've posted and I think you've missed some OOP syntax when calling the super class' constructor.
Try this piece of code and see if it works.

class LinearRegressionModel(nn.Module):
    def __init__(self):
        super().__init__() 
        self.weights = nn.Parameter(torch.randn(1,
            requires_grad=True, 
            dtype=torch.float 
        ))
        self.bias = nn.Parameter(torch.randn(1, 
            requires_grad=True, 
            dtype=torch.float 
        ))