diff options
Diffstat (limited to 'Tasks.org')
| -rw-r--r-- | Tasks.org | 21 |
1 files changed, 14 insertions, 7 deletions
@@ -90,15 +90,22 @@ We also cache primrose results in ~candelabra_cli::candidates~ The CLI should get cost models for all of the candidate types, and for now just print them out. -** DOING Build cost model from benchmark +** DONE Build cost model from benchmark -Fit polynomials for each operation from the benchmarking data. -Possibly helpful crates: - - https://docs.rs/fitme/latest/fitme/fn.fit.html - - https://lib.rs/crates/enterpolation - - https://docs.rs/varpro/latest/varpro/ +We can fit polynomials for each operation from the benchmarking data. +This seems(?) to be working, but could use some further testing. -* BLOCKED Semantic profiler +** TODO Also cache benchmark results + +Currently, we only cache the resulting polynomials, and not the results. +We should probably cache both + +** TODO Fix benchmarks when compiled with release mode + +Currently benchmarks break when compiled in release mode, presumably due to the optimiser being 'too smart'. +Cosmo suggested putting the closures into their own functions. + +* TODO Semantic profiler We need to be able to pick some random candidate type, wrap it with profiling stuff, and run user benchmarks to get data. |