#include <bits/stdc++.h>
#include <atcoder/all>

#include <boost/multiprecision/cpp_dec_float.hpp>
#include <boost/multiprecision/cpp_int.hpp>
#include <boost/rational.hpp>
using Bint = boost::multiprecision::cpp_int;
// 仮数部が10進数で1024桁の浮動小数点数型(TLEしたら小さくする)
using Real = boost::multiprecision::number<boost::multiprecision::cpp_dec_float<1024>>;
using Rat = boost::rational<Bint>;


using namespace std;
using ll = long long;
#define rep(i, s, t) for (ll i = s; i < (ll)(t); i++)
#define rrep(i, s, t) for (ll i = (ll)(t) - 1; i >= (ll)(s); i--)
#define all(x) begin(x), end(x)

#define TT template <typename T>
TT using vec = vector<T>;
TT using minheap = priority_queue<T, vector<T>, greater<T>>;
TT using maxheap = priority_queue<T>;
template <class T1, class T2> bool chmin(T1 &x, T2 y) {
    return x > y ? (x = y, true) : false;
}
template <class T1, class T2> bool chmax(T1 &x, T2 y) {
    return x < y ? (x = y, true) : false;
}
TT int pre_id(vec<T> &A, T x) {  // must sorted
    return lower_bound(all(A), x) - A.begin() - 1;
}
TT int nex_id(vec<T> &A, T x) {  // must sorted
    return upper_bound(all(A), x) - A.begin();
}
struct io_setup {
    io_setup() {
        ios::sync_with_stdio(false);
        std::cin.tie(nullptr);
        cout << fixed << setprecision(15);
    }
} io_setup;
//動的mod : template<int mod> を消して、上の方で変数modを宣言
template<uint32_t mod>
struct modint{
    using mm = modint;
    uint32_t x;
    modint() : x(0) {}
    TT modint(T a=0) : x((a % mod + mod) % mod){}

    friend mm operator+(mm a, mm b) {
        a.x += b.x;
        if(a.x >= mod) a.x -= mod;
        return a;
    }
   friend mm operator-(mm a, mm b) {
        a.x -= b.x;
        if(a.x >= mod) a.x += mod;
        return a;
    }

    //+と-だけで十分な場合、以下は省略して良いです。

    friend mm operator*(mm a, mm b) { return (uint64_t)(a.x) * b.x; }
    friend mm operator/(mm a, mm b) { return a * b.inv(); }
    friend mm& operator+=(mm& a, mm b) { return a = a + b; }
    friend mm& operator-=(mm& a, mm b) { return a = a - b; }
    friend mm& operator*=(mm& a, mm b) { return a = a * b; }
    friend mm& operator/=(mm& a, mm b) { return a = a * b.inv(); }

    mm inv() const {return pow(mod-2);}
    mm pow(const ll& y) const {
        if(!y) return 1;
        mm res = pow(y >> 1);
        res *= res;
        if(y & 1) res *= *this;
        return res;
    }

    friend istream& operator>>(istream &is, mm &a) { 
        ll t;
        cin >> t;
        a = mm(t);
        return is;
    }

    friend ostream& operator<<(ostream &os,  mm a) {
        return os << a.x;
    }

    bool operator==(mm a) {return x == a.x;}
    bool operator!=(mm a) {return x != a.x;}

    bool operator<(const mm& a) const {return x < a.x;}
};

using modint998244353 = modint<998244353>;
using modint1000000007 = modint<1'000'000'007>;

/*
@brief modint
*/
using mint = modint1000000007;

int main() {

    ll n, s, k;
    cin >> n >> s >> k;

    vec<vec<mint>> dp(n + 1, vec<mint>(s + 1, 0));// [i][j] .. i人で総和sを払う場合の和。
    vec<vec<bool>> memo(n + 1, vec<bool>(s + 1, false));


    auto dfs = [&](auto f, ll N, ll S) -> mint {
        if(N == 0) {
            if(S == 0) return 1;
            else return 0;
        }
        if(N < 0 || S < 0) return 0;
        if(memo[N][S]) return dp[N][S];


        //0円が入る場合
        dp[N][S] += f(f, N-1, S - (N-1) * k);

        //0円が入らない場合
        dp[N][S] += f(f, N, S - N);

        memo[N][S] = true;
        return dp[N][S];
    };

    cout << dfs(dfs, n, s) << endl;


}