There are quite a few aspects to the challenge.
First, you'll need to define an order on the buckets. This will allow you to display the visual correctly AND calculate the cumulative % correctly.
One way to do this is to extract the distinct bins to a table. You could construct the table manually (in excel) or experiment in Power Query but what you want is a 2 column table with
Bin Index
0-2 1
3-5 2
and so on
You then create a relationship between the bins(buckets) on both tables.
Next, get a slicer for Month-year.
Then make sure that you can show Qty on a visual with buckets in the correct order. Use 'Sort by Column' feature to sort buckets by the Index.
If you do this succesfully, write a measure to work out the cumulative quantity. Use a table visualisation to test this.
From there, use the measure you've just written to construct a % change measure.
Then put it all together on the Line and Clustered column chart.
If i was you i'd change the test data to something which shows the changes more distinctly because 1,4,5 are very small compared to 667 so the chart will have a large initial value then flatline.
