Before you binned it, it was (sort of) a person's estimate of a typical weekly frequency which is arguably ratio but doesn't exactly fit Stevens' typology of scale.
After you binned it, you have interval -censored data, which retains some ratio information (the highest category is at least 3.5 times the frequency of second category) as long as you don't throw away the known bin endpoints.
Only if you throw away all the binning information do you get down to merely ordinal
That would potentially be losing a lot of information, but it depends on what you want to do with it..