#include <iostream>
#include <vector>
#include <algorithm>
#include <utility>
#include <queue>
#include <tuple>
using namespace std;
typedef tuple<long long, int, int> P;
int dx[4] = {-1, 0, 1, 0}, dy[4] = {0, -1, 0, 1};
int main(){
    int N, M;
    cin >> N >> M;
    vector<int> h(M), w(M);
    vector<long long> c(M);
    vector<vector<long long>> cost(N, vector<long long> (N, 1));
    for(int i = 0; i < M; i++) {
        cin >> h[i] >> w[i] >> c[i];
        cost[h[i] - 1][w[i] - 1] += c[i];
    }
    long long INF = 1e18;
    vector<vector<long long>> d(2, vector<long long>(N * N, INF));
    priority_queue<P, vector<P>, greater<P>> q;
    d[0][0] = 0;
    q.push(P(0, 0, 0));
    while(!q.empty()){
        P v = q.top();
        q.pop();
        long long temp = get<0>(v);
        int s = get<1>(v), t = get<2>(v);
        if(d[s][t] < temp) continue;
        if(s == 0 && cost[t / N][t % N] > 1){
            if(d[1][t] > d[0][t] - cost[t / N][t % N] + 1){
                d[1][t] = d[0][t] - cost[t / N][t % N] + 1;
                q.push(P(d[1][t], 1, t));
            }
        }
        for(int i = 0; i < 4; i++){
            int x = t / N, y = t % N;
            int nx = x + dx[i], ny = y + dy[i];
            if(0 <= nx && nx < N && 0 <= ny && ny < N){
                int nt = nx * N + ny;
                if(d[s][nt] > d[s][t] + cost[nx][ny]){
                    d[s][nt] = d[s][t] + cost[nx][ny];
                    q.push(P(d[s][nt], s, nt));
                }
            }
        }
    }
    cout << min(d[1][N * N - 1], d[0][N * N - 1]) << endl;
}