def call(self, *args, **kwargs) -> Any:
"""
Do not call this directly, use `__call__`:
```
my_lambda(*args, **kwargs)
```
"""
# Unwrap arguments
args = [feature.native for feature in args]
kwargs = {k: kwargs[k].native for k in kwargs}
# call function
result = self.native_module(*args, **kwargs)
# Wrap results
if _isnamedtupleinstance(result):
result_raw = result._asdict()
result_raw = {
k: Tensor(data=result_raw[k], trainable=True)
for k in result_raw
}
return type(result)(**result_raw)
elif isinstance(result, dict):
result = {
k: Tensor(data=result[k], trainable=True)
for k in result
}
elif isinstance(result, list):
result = [Tensor(data=res, trainable=True) for res in result]
else:
result = Tensor(data=result, trainable=True)
return result
def log(tensor: ITensor) -> ITensor:
"""
Compute the logarithm of a tensor.
:param tensor: The tensor of which the log should be computed.
:return: The log tensor.
"""
return Tensor(data=tf.log(tensor.native), trainable=True)
def pow(tensor: ITensor, exponent: ITensor) -> ITensor:
"""
Compute the power of a tensor.
:param tensor: The tensor of which the pow should be computed.
:param exponent: The exponent to take the power with.
:return: The pow tensor.
"""
return Tensor(data=tf.pow(tensor.native, exponent.native), trainable=True)
def clip(tensor: ITensor, min, max) -> ITensor:
"""
Clip a tensor to a value range.
:param tensor: The tensor which should be clipped.
:param min: The minimum value to clip to.
:param max: The maximum value to clip to.
:return: The clipped tensor.
"""
return Tensor(data=torch.clamp(tensor.native, min, max), trainable=True)