Please how to continue.

Can anybody help me how to continue with this game?

trade off pieces and pray

Right off the bat 4...Na5 is bad.  Winning the bishop pair is a great idea, and if white wants to Be2 saving the bishop giving you tempo, but the knight doesn't do much otherwise from a5.  The central dark squares are very important in the Sicilian, and now white dominates them.  I think now would be a great time for 4...Nf6 and you have a little trick, but it's not the main idea, which is to support a ...d5 push.  The trick is 5.e5?!,Ng4 6.Qe2,d5 7.exd6 ep,Bxd6 with development and a firm grip over d5. 

5...a6! is preferred by Bacrot with the idea of bouncing the bishop before playing d5. 

In the game you may have won the bishop pair and saddled white with doubled pawns, but it isn't so simple, that pawn now holds a grip over the important d5 square, so you have no real chance at a d5 break anymore.

7.Ne5 was a mistake by white, it moves the same piece twice and this knight can be challenged with 7...Nc6, also moving the same piece twice but also preparing development.  If 8.Nxc6,bxc6 you have an extra pawn to support the push, but don't bet on white playing this realistically.

7...d6 wasn't a bad move though.

8...Ng6 is good, but Nc6 would watch over the d4 square as well

  13...f5?! gave you a backward pawn center, it's usually bad moving the f and c pawns. 

It just doesn't look right to me, and the position looks to scream 13...Bf6 due to the weak queenside dark squares and forcing white to defend.  13...Bf6 12.Bb2,Bd4 13.Qd2,Qf6 with strong pressure and making a target out of those squares.

18.Nfe4 doesn't do much, 18...Bc6 and after ...Nf4 preparing d5 looks like a great plan as it helps you activate and target white's weaknesses while provoking more. It's an unstoppable plan here and white can't really counter it, and if he tries will just worsten his own position:

18...Bc6 19.Nc3 anticipating Nf4 and a d5 push, 19...Bf6! your bishops become very active and you have excellent prospects of a kingside attack.

18...Nf4 still works via transposition.

20.Nc3?? is a blunder that hands black a wonderful tactic:

20...Nxg2 21.Kxg2,Rxf3 22.Rxf3,Bxf3+ 23.Qxf3,Rxf3 24.Kxf3,Bf6 and you have a passed pawn, strong bishop, and potentially passed kingside pawn for the sac. 25.Bd2 is forced to keep from losing the knight, so black plays 25...Qc6 with the intention of winning the h-pawn and causing mayhem on the kingside.

In the game white is allowing you the Nxg2 combo once more. 

Is this game finished?

I like 21...Nxg2 here, for example 22.Kxg2 Rxf3 23.Rxf3 Rxf3 24.Kg1 Bh4 and white king is in trouble. Black threatens Bf2+ right away with further discovered checks and a strong attack. If needed, the queen can join the rook via f7 - this just looks crushing

I do too, but i wanted to know if it's finished before mentioning it.

The key question everyone should be asking before commenting on an inquiry like this.

So is it?

Yes, it was finished today: http://www.chess.com/echess/game?id=75761880

I can see black went for the sac on g2 but followed differently - maybe even better than Rxf3 but more forcing thus leaving the opponent less opportunities to go wrong ;)

The game is finished. And my opponent couldn't believe it. dzikus has the same idea as me.

And my opponent thought I used an engine. So I post this game to prove that I didn't use any engines.

The puzzle is easy to solve.

ScorpionPawnAttack also has the right idea.

Thanks bean_Fischer, you just outed us all as engine users! 

There are players who hate losing so much that they accuse their opponents of cheating. In online chess, however, finding such a combo is not that hard, provided one does not make the moves instantly but takes the time to analyse and uses the additional board which is allowed. This is my own experience - my opponents sometimes hang their pieces in 14 day game so they certainly do not make use of illegal assistance 

Anyway, this position just shows that tactics-aware players easily find the motive of destroying white's king pawn shelter which is just obvious with those rook battery and sniper bishop on c6. Your opponent cannot just stand he had not predicted that move.

Do not take such ridiculous accusations personally, just enjoy playing chess and improve your tactical vision further 

For those that didn't get the joke, i was kidding about everybody here being engine users.

One can't base engine use solely on a tactic, otherwise everybody is possibly an engine user. Hardly a game goes by without tactics being used.

@Sunshiny: be there no tactics, all games would end in a draw. To win a game one side has to go wrong and the other needs to see and exploit the blunder

That's mostly true. I think it's possible to simply play positionally and win if the opponent makes one or more poor moves.

Spot on, dzikus. I consider myself a positional player. Many of my games ends in >50.

But I also use my tactical weapon when opponent make a weak move. I t maynot be obvious but enough for me to exploit.

The position may seem to be closed, but it just wait for a right move to make it open.

For those who want to be a positional player, they have to prepare for open position that requires tactical play. Opponent can open a semi-closed position at any moment, either by pawn push or sacrifices.

So don't be surprised that suddenly a closed position becomes an open one. I have taken my opponent by surprise. At least I can get 2 pawns and an open King to attack for my knight. And for me that is a good compensation.

Thanks for all who look into my game.