$ stack run --rts-options '+RTS -s'
5000000.0
   2,985,529,488 bytes allocated in the heap
   5,544,085,704 bytes copied during GC
     891,757,168 bytes maximum residency (15 sample(s))
     833,601,864 bytes maximum slop
            2631 MiB total memory in use (0 MB lost due to fragmentation)

                                     Tot time (elapsed)  Avg pause  Max pause
  Gen  0      2625 colls,  2625 par   17.122s   6.625s     0.0025s    0.0261s
  Gen  1        15 colls,    14 par   16.265s   5.922s     0.3948s    1.2558s

  Parallel GC work balance: 8.28% (serial 0%, perfect 100%)

  TASKS: 10 (1 bound, 9 peak workers (9 total), using -N4)

  SPARKS: 0 (0 converted, 0 overflowed, 0 dud, 0 GC'd, 0 fizzled)

  INIT    time    0.001s  (  0.001s elapsed)
  MUT     time    5.771s  (  2.739s elapsed)
  GC      time   33.386s  ( 12.547s elapsed)
  EXIT    time    0.002s  (  0.004s elapsed)
  Total   time   39.161s  ( 15.290s elapsed)

  Alloc rate    517,307,567 bytes per MUT second

  Productivity  14.7% of total user, 17.9% of total elapsed

$ ./.stack-work/dist/x86_64-linux-tinfo6/Cabal-3.2.1.0/build/lhHello0001-exe/lhHello0001-exe +RTS -s
5000000.0
   2,985,526,136 bytes allocated in the heap
   5,581,690,896 bytes copied during GC
     891,757,168 bytes maximum residency (15 sample(s))
   1,072,674,424 bytes maximum slop
            2775 MiB total memory in use (0 MB lost due to fragmentation)

                                     Tot time (elapsed)  Avg pause  Max pause
  Gen  0      2625 colls,  2625 par   16.710s   5.304s     0.0020s    0.0347s
  Gen  1        15 colls,    14 par   21.461s   7.281s     0.4854s    1.6621s

  Parallel GC work balance: 7.25% (serial 0%, perfect 100%)

  TASKS: 10 (1 bound, 9 peak workers (9 total), using -N4)

  SPARKS: 0 (0 converted, 0 overflowed, 0 dud, 0 GC'd, 0 fizzled)

  INIT    time    0.002s  (  0.006s elapsed)
  MUT     time    4.480s  (  2.002s elapsed)
  GC      time   38.171s  ( 12.585s elapsed)
  EXIT    time    0.144s  (  0.008s elapsed)
  Total   time   42.797s  ( 14.600s elapsed)

  Alloc rate    666,341,533 bytes per MUT second

  Productivity  10.5% of total user, 13.7% of total elapsed

{-# LANGUAGE BangPatterns #-}
average' :: [Integer] -> Double
average' xs = f 0 0 xs where
    f ! sum ! count [] = fromIntegral sum / count
    f ! sum ! count (x:xs) = f (sum+x) (count+1) xs

$ stack run --rts-options '+RTS -s'                                                                 
5000000.0
   2,320,166,232 bytes allocated in the heap
       3,932,616 bytes copied during GC
          89,784 bytes maximum residency (2 sample(s))
          56,264 bytes maximum slop
               5 MiB total memory in use (0 MB lost due to fragmentation)

                                     Tot time (elapsed)  Avg pause  Max pause
  Gen  0      2231 colls,  2231 par    0.454s   0.100s     0.0000s    0.0148s
  Gen  1         2 colls,     1 par    0.001s   0.000s     0.0002s    0.0002s

  Parallel GC work balance: 3.28% (serial 0%, perfect 100%)

  TASKS: 10 (1 bound, 9 peak workers (9 total), using -N4)

  SPARKS: 0 (0 converted, 0 overflowed, 0 dud, 0 GC'd, 0 fizzled)

  INIT    time    0.002s  (  0.001s elapsed)
  MUT     time    1.177s  (  0.898s elapsed)
  GC      time    0.455s  (  0.101s elapsed)
  EXIT    time    0.001s  (  0.010s elapsed)
  Total   time    1.636s  (  1.010s elapsed)

  Alloc rate    1,970,541,161 bytes per MUT second

  Productivity  72.0% of total user, 88.9% of total elapsed