Closed kwxm closed 3 months ago
/benchmark nofib
Click here to check the status of your benchmark.
[Time limit = 30s]
Comparing benchmark results of 'nofib' on 'ffe41b864' (base) and 'a6d976575' (PR)
/benchmark nofib
Click here to check the status of your benchmark.
Comparing benchmark results of 'nofib' on 'ffe41b864' (base) and 'a6d976575' (PR)
/benchmark nofib
/benchmark validation
Click here to check the status of your benchmark.
Comparing benchmark results of 'nofib' on 'ffe41b864' (base) and 'a6d976575' (PR)
Click here to check the status of your benchmark.
Comparing benchmark results of 'validation' on 'ffe41b864' (base) and 'a6d976575' (PR)
/benchmark nofib
/benchmark nofib
Click here to check the status of your benchmark.
Comparing benchmark results of 'nofib' on 'ffe41b864' (base) and 'a6d976575' (PR)
Click here to check the status of your benchmark.
Comparing benchmark results of 'nofib' on 'ffe41b864' (base) and 'a6d976575' (PR)
/benchmark nofib
Click here to check the status of your benchmark.
Comparing benchmark results of 'nofib' on 'ffe41b864' (base) and 'a6d976575' (PR)
/benchmark nofib
/benchmark validation
Click here to check the status of your benchmark.
Comparing benchmark results of 'nofib' on 'ffe41b864' (base) and 'a6d976575' (PR)
Click here to check the status of your benchmark.
Comparing benchmark results of 'validation' on 'ffe41b864' (base) and 'a6d976575' (PR)
/benchmark nofib
/benchmark validation
Click here to check the status of your benchmark.
Comparing benchmark results of 'nofib' on 'ffe41b864' (base) and 'a6d976575' (PR)
Click here to check the status of your benchmark.
Comparing benchmark results of 'validation' on 'ffe41b864' (base) and 'a6d976575' (PR)
/benchmark nofib
/benchmark validation
Click here to check the status of your benchmark.
Click here to check the status of your benchmark.
/benchmark validation
/benchmark nofib
Click here to check the status of your benchmark.
Click here to check the status of your benchmark.
This is inconclusive. There's a lot of variation in the results but it's not clear if that's due to the reduction in benchmarking times or inconsistency in the machine.
Also the benchmarks have stopped working because the base commit has vanished, so I'll close this and try again some other time.
This is a followup to #6328. Criterion runs each individual benchmark up to some time limit then analyses the times of the different runs. Our benchmark suites currently get run with different time limits: 60s for
nofib
, 20 forvalidation
, and 15 forlists
. There should be scope for reducing these limits without reducing the accuracy of the results too much. I'm going to do some experiments in this PR to see what happens. Using/benchmark
, the benchmarks will be run in the base branch with the old time limits and in this branch with the new ones, so we should be able to get a decent comparison.