#line 2 "cpplib/util/template.hpp"
#pragma GCC optimize("Ofast")
#pragma GCC optimize("unroll-loops")
#pragma GCC target("avx2")
#include<bits/stdc++.h>
using namespace std;
#line 1 "cpplib/util/ioutil.hpp"
// template <class Head,class... Args>
// std::ostream& output(std::ostream& out,const Head& head,const Args&... args){
//     out>>head;
//     return output(head,args...);
// }
// template <class Head>
// std::ostream& output(std::ostream& out,const Head& head){
//     out>>head;
//     return out;
// }

template<typename T,typename E>
std::ostream& operator<<(std::ostream& out,std::pair<T,E>v){
    out<<"("<<v.first<<","<<v.second<<")";return out;
}

// template <class... Args>
// ostream& operator<<(ostream& out,std::tuple<Args...>v){
//     std::apply(output,v);
//     return out;
// }
#line 8 "cpplib/util/template.hpp"
struct __INIT__{__INIT__(){cin.tie(0);ios::sync_with_stdio(false);cout<<fixed<<setprecision(15);}}__INIT__;
typedef long long lint;
constexpr long long INF=1LL<<60;
constexpr int IINF=1<<30;
constexpr double EPS=1e-10;
#ifndef REACTIVE
    #define endl '\n';
#endif
typedef vector<lint> vec;
typedef vector<vector<lint>> mat;
typedef vector<vector<vector<lint>>> mat3;
typedef vector<string> svec;
typedef vector<vector<string>> smat;
template<typename T>using V=vector<T>;
template<typename T>using VV=V<V<T>>;
template<typename T>inline void output(T t){bool f=0;for(auto i:t){cout<<(f?" ":"")<<i;f=1;}cout<<endl;}
template<typename T>inline void output2(T t){for(auto i:t)output(i);}
template<typename T>inline void debug(T t){bool f=0;for(auto i:t){cerr<<(f?" ":"")<<i;f=1;}cerr<<endl;}
template<typename T>inline void debug2(T t){for(auto i:t)debug(i);}
#define loop(n) for(long long _=0;_<(long long)(n);++_)
#define _overload4(_1,_2,_3,_4,name,...) name
#define __rep(i,a) repi(i,0,a,1)
#define _rep(i,a,b) repi(i,a,b,1)
#define repi(i,a,b,c) for(long long i=(long long)(a);i<(long long)(b);i+=c)
#define rep(...) _overload4(__VA_ARGS__,repi,_rep,__rep)(__VA_ARGS__)
#define _overload3_rev(_1,_2,_3,name,...) name
#define _rep_rev(i,a) repi_rev(i,0,a)
#define repi_rev(i,a,b) for(long long i=(long long)(b)-1;i>=(long long)(a);--i)
#define rrep(...) _overload3_rev(__VA_ARGS__,repi_rev,_rep_rev)(__VA_ARGS__)

#define all(n) begin(n),end(n)
template<typename T,typename E>bool chmin(T& s,const E& t){bool res=s>t;s=min<T>(s,t);return res;}
template<typename T,typename E>bool chmax(T& s,const E& t){bool res=s<t;s=max<T>(s,t);return res;}
const vector<lint> dx={1,0,-1,0,1,1,-1,-1};
const vector<lint> dy={0,1,0,-1,1,-1,1,-1};
#define SUM(v) accumulate(all(v),0LL)
#if __cplusplus>=201703L
    template<typename T,typename ...Args>auto make_vector(T x,int arg,Args ...args){if constexpr(sizeof...(args)==0)return vector<T>(arg,x);else return vector(arg,make_vector<T>(x,args...));}
#endif
#define extrep(v,...) for(auto v:__MAKE_MAT__({__VA_ARGS__}))
#define bit(n,a) ((n>>a)&1)
vector<vector<long long>> __MAKE_MAT__(vector<long long> v){if(v.empty())return vector<vector<long long>>(1,vector<long long>());long long n=v.back();v.pop_back();vector<vector<long long>> ret;vector<vector<long long>> tmp=__MAKE_MAT__(v);for(auto e:tmp)for(long long i=0;i<n;++i){ret.push_back(e);ret.back().push_back(i);}return ret;}
using graph=vector<vector<int>>;
template<typename T>using graph_w=vector<vector<pair<int,T>>>;

#if __cplusplus>=201703L
    constexpr inline long long powll(long long a,long long b){long long res=1;while(b--)res*=a;return res;}
#endif

template<typename T,typename E>pair<T,E>& operator+=(pair<T,E>&s,const pair<T,E>&t){s.first+=t.first;s.second+=t.second;return s;}
template<typename T,typename E>pair<T,E>& operator-=(pair<T,E>&s,const pair<T,E>&t){s.first-=t.first;s.second-=t.second;return s;}
template<typename T,typename E>pair<T,E> operator+(const pair<T,E>&s,const pair<T,E>&t){auto res=s;return res+=t;}
template<typename T,typename E>pair<T,E> operator-(const pair<T,E>&s,const pair<T,E>&t){auto res=s;return res-=t;}
#define BEGIN_STACK_EXTEND(size) void * stack_extend_memory_ = malloc(size);void * stack_extend_origin_memory_;char * stack_extend_dummy_memory_ = (char*)alloca((1+(int)(((long long)stack_extend_memory_)&127))*16);*stack_extend_dummy_memory_ = 0;asm volatile("mov %%rsp, %%rbx\nmov %%rax, %%rsp":"=b"(stack_extend_origin_memory_):"a"((char*)stack_extend_memory_+(size)-1024));
#define END_STACK_EXTEND asm volatile("mov %%rax, %%rsp"::"a"(stack_extend_origin_memory_));free(stack_extend_memory_);
#line 2 "main.cpp"

int main(){
    lint n,k;
    cin>>n>>k;
    lint t=n*(n+1)/2;
    if(t%k){
        cout<<"No"<<endl;
        return 0;
    }
    if(n%k){
        cout<<"No"<<endl;
        return 0;
    }
    t/=k;
    mat ans(k);
    vec su(k);
    priority_queue<pair<lint,lint>,vector<pair<lint,lint>>,greater<pair<lint,lint>>>que;
    rep(i,k){
        que.emplace(0,i);
    }
    if(n/k%2&&k%2&&k*3<=n){
        rrep(i,k*3+1,n+1){
            auto d=que.top();
            que.pop();
            ans[d.second].emplace_back(i);
            que.emplace(d.first+i,d.second);
        }
        mat v(k,vec(3));
        rep(i,k){
            v[i][0]=i+1;
            v[i][1]=k+(i+k/2+1)%k+1;
            v[i][2]=(3*(k*3+1)/2)-v[i][0]-v[i][1];
            ans[i].emplace_back(v[i][0]);
            ans[i].emplace_back(v[i][1]);
            ans[i].emplace_back(v[i][2]);
        }
    }else{
        rrep(i,1,n+1){
            auto d=que.top();
            que.pop();
            ans[d.second].emplace_back(i);
            que.emplace(d.first+i,d.second);
        }
        set<lint>s;
        rep(i,k)if((lint)ans[i].size()!=n/k){
            cout<<"No"<<endl;
            return 0;
        }
        rep(i,k)rep(j,ans[i].size()){
            if(1<=ans[i][j]&&ans[i][j]<=n)s.insert(ans[i][j]);
            else{
                cout<<"No"<<endl;
                return 0;
            }
        }
        rep(i,k)if(SUM(ans[i])!=t){
            cout<<"No"<<endl;
            return 0;
        }
        if((lint)s.size()!=n){
            cout<<"No"<<endl;
            return 0;
        }
    }
    cout<<"Yes"<<endl;
    output2(ans);
}

// int main(){
//     int n,k;
//     cin>>n>>k;
//     vec v(n);
//     rep(i,n)cin>>v[i];
//     sort(all(v));
//     auto get=[&](int to,int from){
//         return (to-from)*v[to-1];
//     };
//     cout<<alien_dp<lint,decltype(get)>(n, k,INF, get)<<endl;
// }

// map<lint,lint> table(lint n){
//     lint l = 1,r;
//     map<lint,lint>res;
//     res[1]=n;
//     while(l <= n){
//         r = n / (n / l) + 1;
//         res[r]=n/r-n/l;
//         l = r;
//     }
//     return res;
// }

// map<lint,lint> merge(map<lint,lint>s,map<lint,lint>t){
//     map<lint,lint>res;
//     for(auto [p,q]:s)res[p]+=q;
//     for(auto [p,q]:t)res[p]-=q;
//     lint pre=0;
//     for(auto [p,q]:res){
//         res[p]+=pre;
//         pre=res[p];
//     }
//     return res;
// }

// map<lint,mint>get_mebius(lint n){
//     Dirichlet<mint>::resize(n);
//     auto dir=Dirichlet<mint>::mebius();
//     vector<mint>suma=dir.a;
//     rep(i,suma.size()-1)suma[i+1]+=suma[i];
//     auto get=[&](long long k){return (k<=Dirichlet<mint>::K?suma[k]:dir.A[n/k]);};
//     lint l = 1,r;
//     map<lint,mint>res;
//     // debug(suma);
//     res[1]=get(1);
//     while(l <= n){
//         r = n / (n / l) + 1;
//         res[r]=get(r);
//         l = r;
//     }
//     return res;
// }

// map<lint,mint> merge2(map<lint,mint>s,map<lint,mint>t){
//     map<lint,mint>res;
//     for(auto [p,q]:s)res[p]=q;
//     for(auto [p,q]:t)res[p]=q;
//     return res;
// }


// void solve(){
//     lint l,r;
//     cin>>l>>r;
//     vec v(r+1);
//     auto tb=merge(table(r),table(l-1));
//     lint pre=0;
//     lint preq=0;
//     auto res=merge2(get_mebius(r),l!=1?get_mebius(l-1):map<lint,mint>());
//     mint ans=0;
//     for(auto [p,q]:tb){
//         // cerr<<p<<" "<<q<<endl;
//         cerr<<p<<" "<<q<<" "<<res[p]<<endl;
//         ans+=mint(2).pow(preq)*(res[pre]-res[p]);
//         pre=p;
//         preq=q;
//     }
//     cout<<ans<<endl;
// }
// int main(){
//     solve();
// }
// int main(){
//     lint t;
//     cin>>t;
//     while(t--)solve();
// }