Claude experiment
#852Comparing
claude-experiment
(2d76dbc
) withreuse-indices
(8fb58d2
)-7%
- Improvement
- 16
- Regression
- 1
- Untouched
- 29
- Skipped
- 190
Different runtime environments detected
Some benchmarks with significant performance changes were compared across different runtime environments, which may affect the accuracy of the results. Learn more about regression causes.Benchmarks
236 total
tests[proof_testing_unify]
benches/ci_benchmarking.rs-7%
25.6 ms27.4 ms
rust_rule_match_overhead[rule_run_1]
benches/rust_api_benchmarking.rs+53%
14 ms9.1 ms
rust_rule_fib[rule_run_1000]
benches/rust_api_benchmarking.rs+35%
340.3 ms251.7 ms
rust_rule_match_overhead[rule_run_1]
benches/rust_api_benchmarking.rs+28%
19.1 ms14.9 ms
rust_rule_fib[rule_run_1000]
benches/rust_api_benchmarking.rs+24%
419 ms338.3 ms
rust_rule_tableaction_hot_path[facts50000_funcs200]
benches/rust_api_benchmarking.rs+19%
35.6 ms29.8 ms
tests[hardboiled_conv1d_32]
benches/ci_benchmarking.rs+17%
1.6 s1.3 s
tests[hardboiled_conv1d_128]
benches/ci_benchmarking.rs+17%
5.1 s4.4 s
tests[hardboiled_conv1d_32]
benches/ci_benchmarking.rs+17%
1.6 s1.4 s
tests[cykjson]
benches/ci_benchmarking.rs+15%
421.3 ms365.8 ms
tests[hardboiled_conv1d_128]
benches/ci_benchmarking.rs+12%
6.1 s5.4 s
rust_rule_tableaction_hot_path[facts50000_funcs200]
benches/rust_api_benchmarking.rs+11%
41.1 ms36.9 ms
tests[python_array_optimize]
benches/ci_benchmarking.rs+8%
5 s4.7 s
tests[cykjson]
benches/ci_benchmarking.rs+8%
542.4 ms501.1 ms
rust_rule_insert_loop[ops1000_funcs200]
benches/rust_api_benchmarking.rs+7%
549.3 µs511.2 µs
tests[python_array_optimize]
benches/ci_benchmarking.rs+7%
6.8 s6.4 s
tests[eggcc-extraction]
benches/ci_benchmarking.rs+6%
1.8 s1.7 s
tests[herbie]
benches/ci_benchmarking.rs+5%
228 ms217.5 ms
tests[herbie]
benches/ci_benchmarking.rs+5%
273.4 ms261.5 ms
rust_rule_insert_loop[ops1000_funcs200]
benches/rust_api_benchmarking.rs+4%
350.4 µs336.7 µs
tests[eggcc-extraction]
benches/ci_benchmarking.rs+4%
1.8 s1.7 s
tests[typeinfer]
benches/ci_benchmarking.rs+3%
223.4 ms215.9 ms
tests[math-microbenchmark]
benches/ci_benchmarking.rs+3%
8.5 s8.3 s
rust_rule_match_with_serialize[rule_run_1]
benches/rust_api_benchmarking.rs+3%
291.4 ms283.7 ms
tests[luminal-llama]
benches/ci_benchmarking.rs+3%
601.2 ms585.5 ms
Commits
Click on a commit to change the comparison rangeBase
reuse-indices
8fb58d2
-7.34%
Exp 70: Replace direct drain_updates_parallel! mid-loop calls with drain_updates!
402d2c6
19 hours ago
by yihozhang+1.93%
Exp 72: Inline leaf-level push_bindings in serial drain to avoid run_plan call
6ee361a
19 hours ago
by yihozhang-0.48%
cleanup
feb4249
17 hours ago
by yihozhang-0.77%
cleanup
741454a
17 hours ago
by yihozhang-0.32%
clean up intersection speedup
5009323
17 hours ago
by yihozhang+1.16%
avoid thread-local storage invocation
92b3fb6
13 hours ago
by yihozhang+91.53%
keep debug assertions
bbcf483
12 hours ago
by yihozhang-91.46%
get_subset can now return a subset reference
d9c9087
12 hours ago
by yihozhang+53.98%
nits
d86f4ec
11 hours ago
by yihozhang-55.71%
cleanup
3156ef9
6 hours ago
by yihozhang+0.72%
nits
2d76dbc
6 hours ago
by yihozhang