Autoscheduler on faster rcnn, stuck on measurement

pytorch 1.7.0 torchvision0.8.1 tvm latest

tuning log: placeholder_red()reduce(x*y)

========== Task 50 (workload key: [“38088ca384bce67161534b1c7aa7b369”, 3]) ========== placeholder = PLACEHOLDER [3] placeholder_red()reduce(x*y)

========== Task 51 (workload key: [“38088ca384bce67161534b1c7aa7b369”, 4]) ========== placeholder = PLACEHOLDER [4] placeholder_red()reduce(x*y)


------------------------------ [ Task Scheduler ]

| ID | Latency (ms) | Speed (GFLOPS) | Trials |

| 0 | - | - | 0 | | 1 | - | - | 0 | | 2 | - | - | 0 | | 3 | - | - | 0 | | 4 | - | - | 0 | | 5 | - | - | 0 | | 6 | - | - | 0 | | 7 | - | - | 0 | | 8 | - | - | 0 | | 9 | - | - | 0 | | 10 | - | - | 0 | | 11 | - | - | 0 | | 12 | - | - | 0 | | 13 | - | - | 0 | | 14 | - | - | 0 | | 15 | - | - | 0 | | 16 | - | - | 0 | | 17 | - | - | 0 | | 18 | - | - | 0 | | 19 | - | - | 0 | | 20 | - | - | 0 | | 21 | - | - | 0 | | 22 | - | - | 0 | | 23 | - | - | 0 | | 24 | - | - | 0 | | 25 | - | - | 0 | | 26 | - | - | 0 | | 27 | - | - | 0 | | 28 | - | - | 0 | | 29 | - | - | 0 | | 30 | - | - | 0 | | 31 | - | - | 0 | | 32 | - | - | 0 | | 33 | - | - | 0 | | 34 | - | - | 0 | | 35 | - | - | 0 | | 36 | - | - | 0 | | 37 | - | - | 0 | | 38 | - | - | 0 | | 39 | - | - | 0 | | 40 | - | - | 0 | | 41 | - | - | 0 | | 42 | - | - | 0 | | 43 | - | - | 0 | | 44 | - | - | 0 | | 45 | - | - | 0 | | 46 | - | - | 0 | | 47 | - | - | 0 | | 48 | - | - | 0 | | 49 | - | - | 0 | | 50 | - | - | 0 | | 51 | - | - | 0 |

Estimated total latency: - ms Trials: 0 Used time : 0 s Next ID: 0

------------------------------ [ Search ]

Generate Sketches #s: 3 Sample Initial Population #s: 1821 fail_ct: 0 Time elapsed: 0.69 GA Iter: 0 Max score: 0.9986 Min score: 0.9288 #Pop: 128 #M+: 0 #M-: 0 GA Iter: 4 Max score: 1.0000 Min score: 0.9872 #Pop: 128 #M+: 1377 #M-: 37 EvolutionarySearch #s: 128 Time elapsed: 2.95

------------------------------ [ Measure ]

Get 64 programs to measure:

The tuning process will not go on in hours.

See my script on how I tune faster rcnn: torchscript-to-tvm/maskrcnn_test.py at master · masahi/torchscript-to-tvm · GitHub

You probably want to use bigger timeout (timeout=100 in my code)