如何加快编译速度

综合讨论区
#1

有如下代码：

import taichi as ti
ti.init(arch=ti.gpu,default_ip=ti.i64)
import time

@ti.func
def is_prime(n):
    结果=True
    if n <= 1:
        结果=False
    elif n == 2:
        结果=True
    elif n % 2 == 0:
        结果=False
    else:
        sqrt = int(ti.sqrt(n))
        for i in range(3, sqrt + 1):
            if n % i == 0:
                结果=False
    return 结果

count=ti.field(dtype=int, shape=())
count[None]=0
@ti.kernel
def prime_count():
    for i in range(2, 10000000001):
        if is_prime(i):
            count[None] += 1

start_time = time.time()
prime_count()
end_time = time.time()
elapsed_time = end_time - start_time
print("Total time: ", elapsed_time, " seconds")
print("Total prime numbers found: ", count)

输出结果

 [Taichi] Starting on arch=cuda

[d:\ProgramData\Miniconda3\envs\taichi\lib\site-packages\taichi\lang\ast\ast_transformer.py:61](file:///D:/ProgramData/Miniconda3/envs/taichi/lib/site-packages/taichi/lang/ast/ast_transformer.py:61): Warning: Casting range_for boundary values from i64 to i32, which may cause numerical issues warnings.warn(

Total time: 0.33608055114746094 seconds Total prime numbers found: 70462980

可以看到，实际运行时间很短，但是在运行前花了两分多钟编译，请问有没有什么设置或写法可以加速编译过程

#2

两分钟不是编译时间是运行时间 :upside_down_face:，你把10000000001改一个小点的数字，会发现一下就算完了