Discuss Shakespeare's deviation from the conventions of the Petrarchan sonnet.

I believe that this question is asking about the structural deviations that Shakespeare's sonnets demonstrate vs. the Petrarchan sonnet format.
Let's start with a similarity and go from there. Both the Shakespearean sonnet and the Petrarchan sonnet are sonnets, which means they are both 14 lines long and written in iambic pentameter. That means each line has 10 syllables in an alternating unstressed/stressed rhythm.
In order to best show how Shakespeare changed the sonnet form, let's look at the traditional Petrarchan sonnet format. This sonnet type divides the poem into two stanzas. The first stanza is eight lines (an octave), and the second stanza is six lines (a sestet). The rhyme scheme of the Petrarchan sonnet is abbaabba, cdecde or cdcdcd. In general, the octave sets up some kind of argument, and the sestet contains the counter argument.
In contrast, the Shakespearean sonnet is made of three quatrains followed by a couplet, and the rhyme scheme is abab cdcd efef gg. In this format, the final couplet is often one of the most important pieces of the entire poem. It usually drives home some kind of conclusion about the previous stanzas. The final couplet is also well known for refuting the previous information of the preceding stanzas, depending on the sonnet in question.