Error Generated when running Sample Program

[Aadit-Ambadkar]

Tested this on a fresh conda environment with just neuralforge installed (and its dependencies) and it generates some errors:

import neuralforge as forge

# Instantiate Tensors:
x = forge.randn((8,4,5))
w = forge.randn((8,5,4), requires_grad = True)
b = forge.randint((4), requires_grad = True)


# Make calculations:
out = x @ w
out += b

# Compute gradients on whole graph:
out.backward()

# Get gradients from specific Tensors:
print(w.grad)
print(b.grad)

Traceback (most recent call last):
  File "/Users/.../autograd_np.py", line 14, in <module>
    out.backward()
  File "/opt/miniconda3/envs/.../neuralforge/tensor_operations.py", line 49, in backward
    self.operation.backward(self.grad, self)
  File "/opt/miniconda3/envs/.../neuralforge/tensor_operations.py", line 273, in backward
    b.backward(db, z)
  File "/opt/miniconda3/envs/.../neuralforge/tensor_operations.py", line 42, in backward
    self.grad += grad
numpy.core._exceptions._UFuncOutputCastingError: Cannot cast ufunc 'add' output from dtype('float64') to dtype('int64') with casting rule 'same_kind'

Not entirely sure how to fix this as it seems like a neuralforge issue and not a my-code issue. I even tried changing the dtype of b._data to float64 (b._data = b._data.astype("float64")) and it still generates the same errors

[eduardoleao052]

Thanks for letting me know! I'll get into it rn