/* #region Head */

// #include <bits/stdc++.h>
#include <algorithm>
#include <array>
#include <bitset>
#include <cassert> // assert.h
#include <cmath>   // math.h
#include <cstring>
#include <ctime>
#include <deque>
#include <fstream>
#include <functional>
#include <iomanip>
#include <iostream>
#include <list>
#include <map>
#include <memory>
#include <numeric>
#include <queue>
#include <random>
#include <set>
#include <sstream>
#include <stack>
#include <string>
#include <unordered_map>
#include <unordered_set>
#include <vector>
using namespace std;

using ll = long long;
using ull = unsigned long long;
using ld = long double;
using pll = pair<ll, ll>;
template <class T> using vc = vector<T>;
template <class T> using vvc = vc<vc<T>>;
using vll = vc<ll>;
using vvll = vvc<ll>;
using vld = vc<ld>;
using vvld = vvc<ld>;
using vs = vc<string>;
using vvs = vvc<string>;
template <class T, class U> using um = unordered_map<T, U>;
template <class T> using pq = priority_queue<T>;
template <class T> using pqa = priority_queue<T, vc<T>, greater<T>>;
template <class T> using us = unordered_set<T>;

#define TREP(T, i, m, n) for (T i = (m), i##_len = (T)(n); i < i##_len; ++(i))
#define TREPM(T, i, m, n) for (T i = (m), i##_max = (T)(n); i <= i##_max; ++(i))
#define TREPR(T, i, m, n) for (T i = (m), i##_min = (T)(n); i >= i##_min; --(i))
#define TREPD(T, i, m, n, d) for (T i = (m), i##_len = (T)(n); i < i##_len; i += (d))
#define TREPMD(T, i, m, n, d) for (T i = (m), i##_max = (T)(n); i <= i##_max; i += (d))

#define REP(i, m, n) for (ll i = (m), i##_len = (ll)(n); i < i##_len; ++(i))
#define REPM(i, m, n) for (ll i = (m), i##_max = (ll)(n); i <= i##_max; ++(i))
#define REPR(i, m, n) for (ll i = (m), i##_min = (ll)(n); i >= i##_min; --(i))
#define REPD(i, m, n, d) for (ll i = (m), i##_len = (ll)(n); i < i##_len; i += (d))
#define REPMD(i, m, n, d) for (ll i = (m), i##_max = (ll)(n); i <= i##_max; i += (d))
#define REPI(itr, ds) for (auto itr = ds.begin(); itr != ds.end(); itr++)
#define REPIR(itr, ds) for (auto itr = ds.rbegin(); itr != ds.rend(); itr++)
#define ALL(x) begin(x), end(x)
#define SIZE(x) ((ll)(x).size())
#define PERM(c)                                                                                                        \
    sort(ALL(c));                                                                                                      \
    for (bool c##p = 1; c##p; c##p = next_permutation(ALL(c)))
#define UNIQ(v) v.erase(unique(ALL(v)), v.end());
#define CEIL(a, b) (((a) + (b)-1) / (b))

#define endl '\n'

constexpr ll INF = 1'010'000'000'000'000'017LL;
constexpr int IINF = 1'000'000'007LL;
constexpr ll MOD = 1'000'000'007LL; // 1e9 + 7
// constexpr ll MOD = 998244353;
constexpr ld EPS = 1e-12;
constexpr ld PI = 3.14159265358979323846;

template <typename T> istream &operator>>(istream &is, vc<T> &vec) { // vector 入力
    for (T &x : vec) is >> x;
    return is;
}
template <typename T> ostream &operator<<(ostream &os, const vc<T> &vec) { // vector 出力 (for dump)
    os << "{";
    REP(i, 0, SIZE(vec)) os << vec[i] << (i == i_len - 1 ? "" : ", ");
    os << "}";
    return os;
}
template <typename T> ostream &operator>>(ostream &os, const vc<T> &vec) { // vector 出力 (inline)
    REP(i, 0, SIZE(vec)) os << vec[i] << (i == i_len - 1 ? "\n" : " ");
    return os;
}

template <typename T, size_t _Nm> istream &operator>>(istream &is, array<T, _Nm> &arr) { // array 入力
    REP(i, 0, SIZE(arr)) is >> arr[i];
    return is;
}
template <typename T, size_t _Nm> ostream &operator<<(ostream &os, const array<T, _Nm> &arr) { // array 出力 (for dump)
    os << "{";
    REP(i, 0, SIZE(arr)) os << arr[i] << (i == i_len - 1 ? "" : ", ");
    os << "}";
    return os;
}

template <typename T, typename U> istream &operator>>(istream &is, pair<T, U> &pair_var) { // pair 入力
    is >> pair_var.first >> pair_var.second;
    return is;
}
template <typename T, typename U> ostream &operator<<(ostream &os, const pair<T, U> &pair_var) { // pair 出力
    os << "(" << pair_var.first << ", " << pair_var.second << ")";
    return os;
}

// map, um, set, us 出力
template <class T> ostream &out_iter(ostream &os, const T &map_var) {
    os << "{";
    REPI(itr, map_var) {
        os << *itr;
        auto itrcp = itr;
        if (++itrcp != map_var.end()) os << ", ";
    }
    return os << "}";
}
template <typename T, typename U> ostream &operator<<(ostream &os, const map<T, U> &map_var) {
    return out_iter(os, map_var);
}
template <typename T, typename U> ostream &operator<<(ostream &os, const um<T, U> &map_var) {
    os << "{";
    REPI(itr, map_var) {
        auto [key, value] = *itr;
        os << "(" << key << ", " << value << ")";
        auto itrcp = itr;
        if (++itrcp != map_var.end()) os << ", ";
    }
    os << "}";
    return os;
}
template <typename T> ostream &operator<<(ostream &os, const set<T> &set_var) { return out_iter(os, set_var); }
template <typename T> ostream &operator<<(ostream &os, const us<T> &set_var) { return out_iter(os, set_var); }
template <typename T> ostream &operator<<(ostream &os, const pq<T> &pq_var) {
    pq<T> pq_cp(pq_var);
    os << "{";
    if (!pq_cp.empty()) {
        os << pq_cp.top(), pq_cp.pop();
        while (!pq_cp.empty()) os << ", " << pq_cp.top(), pq_cp.pop();
    }
    return os << "}";
}

void pprint() { cout << endl; }
template <class Head, class... Tail> void pprint(Head &&head, Tail &&...tail) {
    cout << head;
    if (sizeof...(Tail) > 0) cout << ' ';
    pprint(move(tail)...);
}

// dump
#define DUMPOUT cerr
void dump_func() { DUMPOUT << endl; }
template <class Head, class... Tail> void dump_func(Head &&head, Tail &&...tail) {
    DUMPOUT << head;
    if (sizeof...(Tail) > 0) DUMPOUT << ", ";
    dump_func(move(tail)...);
}

// chmax (更新「される」かもしれない値が前)
template <typename T, typename U, typename Comp = less<>> bool chmax(T &xmax, const U &x, Comp comp = {}) {
    if (comp(xmax, x)) {
        xmax = x;
        return true;
    }
    return false;
}

// chmin (更新「される」かもしれない値が前)
template <typename T, typename U, typename Comp = less<>> bool chmin(T &xmin, const U &x, Comp comp = {}) {
    if (comp(x, xmin)) {
        xmin = x;
        return true;
    }
    return false;
}

// ローカル用
#ifndef ONLINE_JUDGE
#define DEBUG_
#endif

#ifdef DEBUG_
#define DEB
#define dump(...)                                                                                                      \
    DUMPOUT << "  " << string(#__VA_ARGS__) << ": "                                                                    \
            << "[" << to_string(__LINE__) << ":" << __FUNCTION__ << "]" << endl                                        \
            << "    ",                                                                                                 \
        dump_func(__VA_ARGS__)
#else
#define DEB if (false)
#define dump(...)
#endif

#define VAR(type, ...)                                                                                                 \
    type __VA_ARGS__;                                                                                                  \
    cin >> __VA_ARGS__;

template <typename T> istream &operator,(istream &is, T &rhs) { return is >> rhs; }
template <typename T> ostream &operator,(ostream &os, const T &rhs) { return os << ' ' << rhs; }

struct AtCoderInitialize {
    static constexpr int IOS_PREC = 15;
    static constexpr bool AUTOFLUSH = false;
    AtCoderInitialize() {
        ios_base::sync_with_stdio(false), cin.tie(nullptr), cout.tie(nullptr);
        cout << fixed << setprecision(IOS_PREC);
        if (AUTOFLUSH) cout << unitbuf;
    }
} ATCODER_INITIALIZE;

void Yn(bool p) { cout << (p ? "Yes" : "No") << endl; }
void YN(bool p) { cout << (p ? "YES" : "NO") << endl; }

/* #endregion */

// #include <atcoder/all>
// using namespace atcoder;

/* #region Graph */

// エッジ（本来エッジは双方向だが，ここでは単方向で管理）
template <class weight_t = int, class flow_t = int> struct Edge {
    ll src;          // エッジ始点となる頂点
    ll dst;          // エッジ終点となる頂点
    weight_t weight; // 重み
    flow_t cap;
    Edge() : src(0), dst(0), weight(0) {}
    Edge(ll src, ll dst, weight_t weight) : src(src), dst(dst), weight(weight) {}
    Edge(ll src, ll dst, weight_t weight, flow_t cap) : src(src), dst(dst), weight(weight), cap(cap) {}
    // Edge 標準出力
    friend ostream &operator<<(ostream &os, Edge &edge) {
        os << "(" << edge.src << " -> " << edge.dst << ", " << edge.weight << ")";
        return os;
    }
};
// 同じ頂点を始点とするエッジ集合
template <class weight_t = int, class flow_t = int> class Node : public vc<Edge<weight_t, flow_t>> {
  public:
    ll idx;
    Node() : vc<Edge<weight_t, flow_t>>() {}
    // void add(int a, int b, weight_t w, flow_t cap) { this->emplace_back(a, b, w, cap); };
};
// graph[i] := 頂点 i を始点とするエッジ集合
template <class weight_t = int, class flow_t = int> class Graph : public vc<Node<weight_t, flow_t>> {
  public:
    Graph() : vc<Node<weight_t, flow_t>>() {}
    Graph(int n) : vc<Node<weight_t, flow_t>>(n) { REP(i, 0, n)(*this)[i].idx = i; }
    // 単方向
    void add_arc(int a, int b, weight_t w = 1, flow_t cap = 1) { (*this)[a].emplace_back(a, b, w, cap); }
    // 双方向
    void add_edge(int a, int b, weight_t w = 1, flow_t cap = 1) { add_arc(a, b, w, cap), add_arc(b, a, w, cap); }
};
// using Array = vc<Weight>;
// using Matrix = vc<Array>;

/* #endregion */

/* #region Dinic */

// 最大流 (Dinic) | libalgo https://tubo28.me/compprog/algorithm/dinic/
template <class weight_t, class flow_t> struct dinic {
    int n, s, t;
    vc<int> level, prog, que;
    vc<vc<flow_t>> cap, flow;
    vc<vc<int>> g;
    flow_t inf;

    dinic(const int n)
        : n(n), cap(n, vc<flow_t>(n)), flow(n, vc<flow_t>(n)), g(n, vc<int>()),
          inf(std::numeric_limits<flow_t>::max() / 8) {}
    dinic(const Graph<weight_t, flow_t> &graph)
        : n(graph.size()), cap(n, vc<flow_t>(n)), flow(n, vc<flow_t>(n)), g(n, vc<int>()),
          inf(std::numeric_limits<flow_t>::max() / 8) {
        build(graph);
    }
    void build(const Graph<weight_t, flow_t> &graph) {
        for (int i = 0; i < n; i++) {
            for (auto &e : graph[i]) {
                int u = e.src, v = e.dst;
                flow_t c = e.cap;
                cap[u][v] += c;
                cap[v][u] += c;
                flow[v][u] += c;
                g[u].push_back(v);
                g[v].push_back(u);
            }
        }
    }
    inline flow_t residue(int u, int v) { return cap[u][v] - flow[u][v]; }
    flow_t solve(int s_, int t_) {
        this->t = t_, this->s = s_;
        que.resize(n + 1);
        flow_t res = 0;
        while (levelize()) {
            prog.assign(n, 0);
            res += augment(s, inf);
        }
        return res;
    }
    bool levelize() {
        int l = 0, r = 0;
        level.assign(n, -1);
        level[s] = 0;
        que[r++] = s;
        while (l != r) {
            int v = que[l++];
            if (v == t) break;
            for (const int &d : g[v])
                if (level[d] == -1 && residue(v, d) != 0) {
                    level[d] = level[v] + 1;
                    que[r++] = d;
                }
        }
        return level[t] != -1;
    }
    flow_t augment(int v, flow_t lim) {
        flow_t res = 0;
        if (v == t) return lim;
        for (int &i = prog[v]; i < (int)g[v].size(); i++) {
            const int &d = g[v][i];
            if (residue(v, d) == 0 || level[v] >= level[d]) continue;
            const flow_t aug = augment(d, std::min(lim, residue(v, d)));
            flow[v][d] += aug;
            flow[d][v] -= aug;
            res += aug;
            lim -= aug;
            if (lim == 0) break;
        }
        return res;
    }
};

/* #endregion */

// Problem
void solve() {
    VAR(ll, n, m); //
    vll k(n), c(n);
    vvll a(n), b(n);
    REP(i, 0, n) {
        cin >> k[i], c[i]; //
        a[i].resize(k[i]);
        b[i].resize(k[i]);
        cin >> a[i], b[i];
        REP(j, 0, k[i]) a[i][j]--;
    }

    // 各科目の 勉強する/しない，の選択をする，燃やす埋める問題に帰着する
    // ll ma = *max_element(ALL(c));
    // ma = max(m, ma);
    // ll offset = ma;

    ll sum = m * n;
    REP(i, 0, n) sum += accumulate(ALL(b[i]), 0LL);

    using Flow = ll;
    Graph<int, Flow> g(n + 2);
    ll S = n, T = n + 1;
    REP(i, 0, n) {
        g.add_arc(S, i, 0, c[i]);                           // 勉強するとき
        g.add_arc(i, T, 0, m + accumulate(ALL(b[i]), 0LL)); // 勉強しないとき
        REP(j, 0, k[i]) {
            g.add_arc(a[i][j], i, 0, b[i][j]); //
        }
    }
    dinic<int, Flow> dn(g);
    // f = 最小費用
    Flow f = dn.solve(S, T);
    // dump(sum, f);
    pprint(sum - f);
}

// entry point
int main() {
    solve();
    return 0;
}