There is at least one possible error here (係 v.s. 喺), so I'll assume a nonzero possibility of characters being misassigned. In that case, I'll assume the pronunciation is correct except for the tone.
I have never heard gam3
(generally written as 咁) used (correctly) outside of a context where it comes before an adjective as a means of indicating the extent something is <adjective>
.
Conversely, there is also gam2
, which has the same pronunciation apart from the tone. (The dictionary gives as 噉.) Use of this gam2
alone (i.e. not in some sort of set phrase like 噉樣) indicates something like "in this case" (e.g. 今日落雨,噉我哋唔好出去啦---it is raining today, so let's not go out). (The dictionary also suggests 恁 as a particle with the romanisation gam
with a meaning that could fit, but the use-case appears identical to 噉, so I cannot differentiate between them.)
It is worth noting that people seem to sometimes conflate the two---e.g. 咁樣 and 噉樣 both appear in the Google IME. In fact, 咁嘅 is a lot more common than 噉嘅, judging from Google search and IME suggestions, even though gam3
is definitely not the right tone. Thus, it looks like there's some variation in what the correct transcription for a character not used formally is. (I later found similar ambiguities when I tried checking the character for zung6
.)
Regardless of what the correct character is, gam2
makes more sense, because I can group it with 係 in a way that makes sense to me. This means that the translation should be something like:
I still haven't lost hope, so I'm waiting here.
For what it's worth, I'm having a lot of trouble catching any instance of gam
in the audio of the song ("想你想得好孤寂") in question. If the particle were actually absent, though, the line would not feel complete. (I would interpret it as either I still haven't lost hope, waiting here or I still haven't lost hope that waiting here . . .)
gam
seems to have been omitted or afforded very little stress.)hai
here should be "at" (which I suppose you may have alluded to in your comment). I'm guessing that the romanisation is based on the characters, rather than the actual song audio?