Hello!
In def init Encoder and Decoder must be with 2 arguments (vocab_size, units)?
In part “call”
context, target = to_translate, sr_translation -?
TypeError: Exception encountered when calling layer ‘translator_18’ (type Translator).
Encoder.init() missing 1 required positional argument: ‘units’
Call arguments received by layer ‘translator_18’ (type Translator):
• inputs=(‘tf.Tensor(shape=(64, 14), dtype=int64)’, ‘tf.Tensor(shape=(64, 15), dtype=int64)’)
This is right!
This you should refer to this comment:
Args:
inputs (tuple(tf.Tensor, tf.Tensor)): Tuple containing the context (sentence to translate) and the target (shifted-to-the-right
I am not sure that I understand this right…
To convert “to_translate” ( now <tf.Tensor: shape=(64, 14), dtype=int64…) in other format as variables?
Inputs is a tuple with 2 values inputs = (value1, value2)
Hello!
with “context, target = inputs(tuple(to_translate)), inputs(tuple(sr_translation))”
also with 'context, target = inputs(tuple(to_translate),tuple(sr_translation))"
Type error
Type error
What else may be wrong?
—> 38 context, target = inputs(tuple(to_translate)), inputs(tuple(sr_translation)) # In this case inputs is a tuple consisting of the context and the target, unpack it into single variables
39
40 # Pass the context through the encoder
TypeError: Exception encountered when calling layer ‘translator_7’ (type Translator).
‘tuple’ object is not callable
Call arguments received by layer ‘translator_7’ (type Translator):
• inputs=(‘tf.Tensor(shape=(64, 14), dtype=int64)’, ‘tf.Tensor(shape=(64, 15), dtype=int64)’)
TypeError: Exception encountered when calling layer ‘translator_8’ (type Translator).
‘tuple’ object is not callable
Call arguments received by layer ‘translator_8’ (type Translator):
• inputs=(‘tf.Tensor(shape=(64, 14), dtype=int64)’, ‘tf.Tensor(shape=(64, 15), dtype=int64)’)
You’re overthinking, it’s simpler:
context, target = inputs
Note the docstring:
Args: inputs (tuple(tf.Tensor, tf.Tensor)): Tuple containing the context (sentence to translate) and the target (shifted-to-the-right translation)
Cheers
Thank you! You saved me :))) 