I would recommend sitting down and trying to teach yourself intuitive f2l. If you know the general idea of it, and can understand how your beginners method first layer corners, and middle layer edges algs work, then you should be able to figure out some basic f2l cases on your own. Start with a solved cube, remove an f2l pair and put it in the U layer (I'm assuming you have cross on bottom at this point.) Now break apart the pair in some way and observe how the corner and edge move. Try to reassemble the pair and reinsert it. Try to figure out as much of f2l as you can on your own, when you get stuck/confused go check out f2l tutorials from badmephisto, erik akkersdijk, thrawst, rowe hessler (rowehess on youtube), and other cubers. You can also go straight to the wiki's f2l page - [wiki]f2l[/wiki] to learn algs for difficult cases you may run into.
As for last layer, what method are you using now? I'm assuming it's some variation of 4-look last layer where you orient edges, oriented corners, permute corners and permute edges individually using 1 or maybe 2 repetitions of an algorithm. Or do you use the "Dan Brown" method (rubik's booklet method)?
I'm probably going to tell you to finish full 4 Look Last Layer (17 algs total, 3 for EO, 7 for CO, 3 for CP and 4 for EP)
when you finish that and start to see improvement, finish PLL (a total of 21 algs, 7 of which you would know from 4 Look last layer)
Then finish OLL, 47 more algorithms, but I swear it's really not that hard to learn. Much easier than PLL (that is if you learn full PLL first) because so many algs are very similar and most are just combinations of simple triggers, such as sexy move (R U R' U'), sledgehammer (R' F R F') sune (R U R' U R U2 R') and others.
Take your time, as much as you feel you need between learning algs. And of course don't forget that any time you start to dislike the method you are using you can learn a new one such as [wiki]petrus[/wiki] , [wiki]roux[/wiki] , [wiki]zz[/wiki] and [wiki]heise[/wiki]
