N,I = map(int,input().split())

inf = 10 ** 10
dp = [-inf] * (I + 1)
dp[0] = 0
for _ in range(N):
    s,a = map(int,input().split())
    for i in range(I,-1,-1):
        if i + s <= I:
            dp[i + s] = max(dp[i + s],dp[i] + a)
print(max(dp))