// ---------- begin ModInt ----------
mod modint {

    #[allow(dead_code)]
    pub struct Mod;
    impl ConstantModulo for Mod {
        const MOD: u32 = 1_000_000_007;
    }

    #[allow(dead_code)]
    pub struct StaticMod;
    static mut STATIC_MOD: u32 = 0;
    impl Modulo for StaticMod {
        fn modulo() -> u32 {
            unsafe { STATIC_MOD }
        }
    }

    #[allow(dead_code)]
    impl StaticMod {
        pub fn set_modulo(p: u32) {
            unsafe {
                STATIC_MOD = p;
            }
        }
    }

    use std::marker::*;
    use std::ops::*;

    pub trait Modulo {
        fn modulo() -> u32;
    }

    pub trait ConstantModulo {
        const MOD: u32;
    }

    impl<T> Modulo for T
    where
        T: ConstantModulo,
    {
        fn modulo() -> u32 {
            T::MOD
        }
    }

    pub struct ModInt<T>(pub u32, PhantomData<T>);

    impl<T> Clone for ModInt<T> {
        fn clone(&self) -> Self {
            ModInt::new_unchecked(self.0)
        }
    }

    impl<T> Copy for ModInt<T> {}

    impl<T: Modulo> Add for ModInt<T> {
        type Output = ModInt<T>;
        fn add(self, rhs: Self) -> Self::Output {
            let mut d = self.0 + rhs.0;
            if d >= T::modulo() {
                d -= T::modulo();
            }
            ModInt::new_unchecked(d)
        }
    }

    impl<T: Modulo> AddAssign for ModInt<T> {
        fn add_assign(&mut self, rhs: Self) {
            *self = *self + rhs;
        }
    }

    impl<T: Modulo> Sub for ModInt<T> {
        type Output = ModInt<T>;
        fn sub(self, rhs: Self) -> Self::Output {
            let mut d = self.0 - rhs.0;
            if d >= T::modulo() {
                d += T::modulo();
            }
            ModInt::new_unchecked(d)
        }
    }

    impl<T: Modulo> SubAssign for ModInt<T> {
        fn sub_assign(&mut self, rhs: Self) {
            *self = *self - rhs;
        }
    }

    impl<T: Modulo> Mul for ModInt<T> {
        type Output = ModInt<T>;
        fn mul(self, rhs: Self) -> Self::Output {
            let v = self.0 as u64 * rhs.0 as u64 % T::modulo() as u64;
            ModInt::new_unchecked(v as u32)
        }
    }

    impl<T: Modulo> MulAssign for ModInt<T> {
        fn mul_assign(&mut self, rhs: Self) {
            *self = *self * rhs;
        }
    }

    impl<T: Modulo> Neg for ModInt<T> {
        type Output = ModInt<T>;
        fn neg(self) -> Self::Output {
            if self.0 == 0 {
                Self::zero()
            } else {
                Self::new_unchecked(T::modulo() - self.0)
            }
        }
    }

    impl<T> std::fmt::Display for ModInt<T> {
        fn fmt<'a>(&self, f: &mut std::fmt::Formatter<'a>) -> std::fmt::Result {
            write!(f, "{}", self.0)
        }
    }

    impl<T> std::fmt::Debug for ModInt<T> {
        fn fmt<'a>(&self, f: &mut std::fmt::Formatter<'a>) -> std::fmt::Result {
            write!(f, "{}", self.0)
        }
    }

    impl<T: Modulo> std::str::FromStr for ModInt<T> {
        type Err = std::num::ParseIntError;
        fn from_str(s: &str) -> Result<Self, Self::Err> {
            let val = s.parse::<u32>()?;
            Ok(ModInt::new(val))
        }
    }

    impl<T: Modulo> From<usize> for ModInt<T> {
        fn from(val: usize) -> ModInt<T> {
            ModInt::new_unchecked((val % T::modulo() as usize) as u32)
        }
    }

    impl<T: Modulo> From<u64> for ModInt<T> {
        fn from(val: u64) -> ModInt<T> {
            ModInt::new_unchecked((val % T::modulo() as u64) as u32)
        }
    }

    impl<T: Modulo> From<i64> for ModInt<T> {
        fn from(val: i64) -> ModInt<T> {
            let m = T::modulo() as i64;
            ModInt::new((val % m + m) as u32)
        }
    }

    #[allow(dead_code)]
    impl<T> ModInt<T> {
        pub fn new_unchecked(d: u32) -> Self {
            ModInt(d, PhantomData)
        }
        pub fn zero() -> Self {
            ModInt::new_unchecked(0)
        }
        pub fn one() -> Self {
            ModInt::new_unchecked(1)
        }
        pub fn is_zero(&self) -> bool {
            self.0 == 0
        }
    }

    #[allow(dead_code)]
    impl<T: Modulo> ModInt<T> {
        pub fn new(d: u32) -> Self {
            ModInt::new_unchecked(d % T::modulo())
        }
        pub fn pow(&self, mut n: u64) -> Self {
            let mut t = Self::one();
            let mut s = *self;
            while n > 0 {
                if n & 1 == 1 {
                    t *= s;
                }
                s *= s;
                n >>= 1;
            }
            t
        }
        pub fn inv(&self) -> Self {
            assert!(self.0 != 0);
            self.pow(T::modulo() as u64 - 2)
        }
    }
}
// ---------- end ModInt ----------
// ---------- begin Precalc ----------
mod precalc {
    use super::modint::*;
    #[allow(dead_code)]
    pub struct Precalc<T> {
        inv: Vec<ModInt<T>>,
        fact: Vec<ModInt<T>>,
        ifact: Vec<ModInt<T>>,
    }
    #[allow(dead_code)]
    impl<T: Modulo> Precalc<T> {
        pub fn new(n: usize) -> Precalc<T> {
            let mut inv = vec![ModInt::one(); n + 1];
            let mut fact = vec![ModInt::one(); n + 1];
            let mut ifact = vec![ModInt::one(); n + 1];
            for i in 2..(n + 1) {
                fact[i] = fact[i - 1] * ModInt::new_unchecked(i as u32);
            }
            ifact[n] = fact[n].inv();
            if n > 0 {
                inv[n] = ifact[n] * fact[n - 1];
            }
            for i in (1..n).rev() {
                ifact[i] = ifact[i + 1] * ModInt::new_unchecked((i + 1) as u32);
                inv[i] = ifact[i] * fact[i - 1];
            }
            Precalc {
                inv: inv,
                fact: fact,
                ifact: ifact,
            }
        }
        pub fn inv(&self, n: usize) -> ModInt<T> {
            assert!(n > 0);
            self.inv[n]
        }
        pub fn fact(&self, n: usize) -> ModInt<T> {
            self.fact[n]
        }
        pub fn ifact(&self, n: usize) -> ModInt<T> {
            self.ifact[n]
        }
        pub fn perm(&self, n: usize, k: usize) -> ModInt<T> {
            if k > n {
                return ModInt::zero();
            }
            self.fact[n] * self.ifact[n - k]
        }
        pub fn comb(&self, n: usize, k: usize) -> ModInt<T> {
            if k > n {
                return ModInt::zero();
            }
            self.fact[n] * self.ifact[k] * self.ifact[n - k]
        }
    }
}
// ---------- end Precalc ----------

use modint::*;
type M = ModInt<Mod>;
// ---------- begin karatsuba multiplication ----------
fn karatsuba<T>(a: &[T], b: &[T], c: &mut [T], zero: T, buf: &mut [T])
where T: std::marker::Copy +
         std::ops::Add<Output = T> +
         std::ops::Sub<Output = T> +
         std::ops::Mul<Output = T> +
{
    assert!(a.len() == b.len());
    let n = a.len();
    if n <= 16 {
        for (i, a) in a.iter().enumerate() {
            for (c, b) in c[i..].iter_mut().zip(b) {
                *c = *c + *a * *b;
            }
        }
        return;
    }
    if n & 1 == 1 {
        karatsuba(&a[1..], &b[1..], &mut c[2..], zero, buf);
        let x = a[0];
        let y = b[0];
        c[0] = c[0] + x * y;
        for ((c, a), b) in c[1..].iter_mut().zip(&a[1..]).zip(&b[1..]) {
            *c = *c + x * *b + *a * y;
        }
        return;
    }
    let m = n / 2;
    let (fa, ta) = a.split_at(m);
    let (fb, tb) = b.split_at(m);
    karatsuba(fa, fb, &mut c[..n], zero, buf);
    karatsuba(ta, tb, &mut c[n..], zero, buf);
    let (x, buf) = buf.split_at_mut(m);
    let (y, buf) = buf.split_at_mut(m);
    let (z, buf) = buf.split_at_mut(n);
    z.iter_mut().for_each(|z| *z = zero);
    let xpq = x.iter_mut().zip(fa).zip(ta);
    let yrs = y.iter_mut().zip(fb).zip(tb);
    for (((x, p), q), ((y, r), s)) in xpq.zip(yrs) {
        *x = *p + *q;
        *y = *r + *s;
    }
    karatsuba(x, y, z, zero, buf);
    for ((z, p), q) in z.iter_mut().zip(&c[..n]).zip(&c[n..]) {
        *z = *z - (*p + *q);
    }
    for (c, z) in c[m..].iter_mut().zip(z) {
        *c = *c + *z;
    }
}

pub fn multiply<T>(a: &[T], b: &[T], zero: T) -> Vec<T>
where T: std::marker::Copy +
         std::ops::Add<Output = T> +
         std::ops::Sub<Output = T> +
         std::ops::Mul<Output = T> +
{
    assert!(!a.is_empty() && !b.is_empty());
    let mut i = 0;
    let mut j = 0;
    let mut ans = vec![zero; a.len() + b.len() - 1];
    let mut buf = vec![zero; 4 * a.len().min(b.len())];
    let mut c = Vec::with_capacity(2 * a.len().min(b.len()));
    while i < a.len() && j < b.len() {
        let x = a.len() - i;
        let y = b.len() - j;
        let z = x.min(y);
        c.clear();
        c.resize(2 * z, zero);
        karatsuba(&a[i..(i + z)], &b[j..(j + z)], &mut c, zero, &mut buf);
        for (ans, c) in ans[(i + j)..].iter_mut().zip(c.iter()) {
            *ans = *ans + *c;
        }
        if x <= y {
            j += z;
        } else {
            i += z;
        }
    }
    ans.truncate(a.len() + b.len() - 1);
    ans
}
// ---------- end karatsuba multiplication ----------

fn inverse(a: &[M], n: usize) -> Vec<M> {
    assert!(a[0].0 > 0);
    let mut b = vec![a[0].inv(); 1];
    let mut k = 1;
    while k < n {
        k *= 2;
        let c = multiply(&b, &b, M::zero());
        let x = a.iter().take(k).cloned().collect::<Vec<_>>();
        let y = multiply(&x, &c, M::zero());
        b.resize(k, ModInt::zero());
        for (b, y) in b.iter_mut().zip(y) {
            *b = *b + *b - y;
        }
    }
    b
}

fn run() {
    let mut s = String::new();
    use std::io::*;
    std::io::stdin().read_to_string(&mut s).unwrap();
    let mut it = s.trim().split_whitespace();
    let k: usize = it.next().unwrap().parse().unwrap();
    let n: usize = it.next().unwrap().parse().unwrap();
    let mut a = vec![M::zero(); 100_000 + 1];
    a[0] = M::one();
    for _ in 0..n {
        let x: usize = it.next().unwrap().parse().unwrap();
        a[x] = -ModInt::one();
    }
    let inv = inverse(&a, k + 1);
    println!("{}", inv[k].0);
}

fn main() {
    run();
}