ggml-org/whisper.cpp
Should we use process time for benchmarks/timings?
Aperta
#382 aperta il 7 gen 2023
enhancementgood first issue
Metriche repository
- Star
- (52.942 stelle)
- Metriche merge PR
- (Merge medio 5g 20h) (36 PR mergiate in 30 g)
Descrizione
Currently, the bench tool simply spits out the timings from whisper_print_timings. These timings are not process specific and as such are influenced by other processes on the system.
Should we use something like CLOCK_PROCESS_CPUTIME_ID instead of CLOCK_MONOTONIC in ggml_time_*? This would make the benchmarks in #89 outdated, but would make future benchmarks more realistic.