Even for the context provided in Assignment itself it is showing wrong answer and I tried with my own context and it is showing wrong answers. What to do in this case?
Can you refine your question? Do you have trouble passing the Assignment?
Hi @arvyzukai , I am having the same problem as the original poster here and I got the same answer (June 1840 instead of 15 December 1840). (though I did get the January 9, 1957 answer correct)
I passed the Unit Test but when I submit, I get the following error:
There was an error grading your submission. Details:
Exception encountered when calling layer ‘multi_head_attention_3’ (type MultiHeadAttention).
{{function_node _wrapped__Einsum_N_2_device/job:localhost/replica:0/task:0/device:CPU:0}} Expected dimension 1 at axis 0 of the input shaped [2,1,2,128] but got dimension 2 [Op:Einsum] name:
Call arguments received by layer ‘multi_head_attention_3’ (type MultiHeadAttention):
• query=tf.Tensor(shape=(2, 1, 128), dtype=float32)
• value=tf.Tensor(shape=(1, 13, 128), dtype=float32)
• key=tf.Tensor(shape=(1, 13, 128), dtype=float32)
• attention_mask=tf.Tensor(shape=(1, 1, 13), dtype=float32)
• return_attention_scores=True
• training=False
• use_causal_mask=False