# Scales dot product attention **URL:** \ **Category:** Sequence Models\ **Tags:** coursera-platform\ **Created:** [June 17, 2021, 12:25pm UTC](https://community.deeplearning.ai/t/scales-dot-product-attention/14712 "2021-06-17T12:25:49Z")\ **Posts on this page:** 3\ **Page:** 1
**Author:** ![isuru](https://avatars.discourse-cdn.com/v4/letter/i/ecb155/32.png) [@isuru](https://community.deeplearning.ai/u/isuru)\ **Post date:** [June 17, 2021, 12:25pm UTC](https://community.deeplearning.ai/t/scales-dot-product-attention/14712/1 "2021-06-17T12:25:49Z")
in scaled\_dot\_product\_attention\_test(target) 10 assert np.allclose(weights, [[0.2589478, 0.42693272, 0.15705977, 0.15705977], 11 [0.2772748, 0.2772748, 0.2772748, 0.16817567], —\> 12 [0.33620113, 0.33620113, 0.12368149, 0.2039163]]) I got the error like this please help me I am stuck at this place for 4 hours now if mask is not None: scaled\_attention\_logits += mask \* -1e9 is the above statement is write please help I used-\> dk = k.shape[-1] scaled\_attention\_logits = matmul\_qk/np.sqrt(dk) also used tf.keras.activations.softmax(scaled\_attention\_logits, axis = dk) to get attention weights ---
**Author:** ![MichaelAnderson](https://avatars.discourse-cdn.com/v4/letter/m/b2d939/32.png) [@MichaelAnderson](https://community.deeplearning.ai/u/MichaelAnderson)\ **Post date:** [June 17, 2021, 4:12pm UTC](https://community.deeplearning.ai/t/scales-dot-product-attention/14712/2 "2021-06-17T16:12:34Z")
Hi @isuru use tf.linalg.matmul for the output, that should fix it ---
**Author:** ![TMosh](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/tmosh/32/9321_2.png) [@TMosh](https://community.deeplearning.ai/u/TMosh)\ **Post date:** [June 18, 2021, 12:11am UTC](https://community.deeplearning.ai/t/scales-dot-product-attention/14712/3 "2021-06-18T00:11:13Z")
@isuru I do not see any issues there. Please edit your message and remove your code fragments. Posting your code breaks the Honor Code.