215-exploratory-refactor-encoderfile_core---encoderfile - Branch - mozilla-ai/encoderfile

feat: refactor encoderfile core -> encoderfile

#216Merged

Comparing

215-exploratory-refactor-encoderfile_core---encoderfile

(

ce7dd3e

) with

main

(

89ae444

)

-24%

Improvement: 1

Regression: 3

Untouched: 16

Skipped: 20

Benchmarks

40 total

sequence_classification_postprocess[16]

encoderfile/benches/postprocessing.rs

+10%

33.6 µs*30.5 µs

bench_tok_cls_softmax[(16, 16, 2)]

encoderfile/benches/benchmark_transforms.rs

-19%

149.2 µs*184 µs

bench_tok_cls_softmax[(128, 256, 32)]

encoderfile/benches/benchmark_transforms.rs

-22%

27.9 ms*35.7 ms

bench_tok_cls_softmax[(32, 128, 8)]

encoderfile/benches/benchmark_transforms.rs

-24%

1.4 ms*1.8 ms

bench_l2_normalize[Axis(1)]

encoderfile/benches/l2_norm.rs

+1%

3 ms*3 ms

token_classification_postprocess[(8, 16)]

encoderfile/benches/postprocessing.rs

3.6 ms*3.6 ms

token_classification_postprocess[(64, 512)]

encoderfile/benches/postprocessing.rs

61.2 ms*61.2 ms

embedding_postprocess[(8, 16, 384)]

encoderfile/benches/postprocessing.rs

4.6 ms*4.6 ms

bench_embedding_l2_normalization[(16, 16, 16)]

encoderfile/benches/benchmark_transforms.rs

337 µs*337 µs

token_classification_postprocess[(16, 128)]

encoderfile/benches/postprocessing.rs

8.2 ms*8.2 ms

embedding_postprocess[(16, 128, 768)]

encoderfile/benches/postprocessing.rs

25.2 ms*25.2 ms

embedding_postprocess[(64, 512, 1024)]

encoderfile/benches/postprocessing.rs

311.3 ms*312.1 ms

bench_embedding_l2_normalization[(32, 128, 384)]

encoderfile/benches/benchmark_transforms.rs

65.1 ms*65.3 ms

bench_embedding_l2_normalization[(32, 256, 768)]

encoderfile/benches/benchmark_transforms.rs

266.1 ms*267 ms

sequence_classification_postprocess[8]

encoderfile/benches/postprocessing.rs

-1%

11 µs*11.1 µs

sequence_classification_postprocess[64]

encoderfile/benches/postprocessing.rs

-1%

47.8 µs*48.5 µs

bench_l2_normalize[Axis(0)]

encoderfile/benches/l2_norm.rs

-1%

1.5 ms*1.6 ms

bench_seq_cls_softmax[(16, 2)]

encoderfile/benches/benchmark_transforms.rs

-7%

127.5 µs*137 µs

bench_seq_cls_softmax[(128, 32)]

encoderfile/benches/benchmark_transforms.rs

-8%

269.5 µs*292.3 µs

bench_seq_cls_softmax[(32, 8)]

encoderfile/benches/benchmark_transforms.rs

-8%

134.7 µs*147 µs

The benchmarks below were skipped, so their baseline results are used instead. If they were deleted in your codebase, archive them to remove them from the performance reports.
Learn more about archiving benchmarks

bench_seq_cls_softmax[(32, 8)]

encoderfile-core/benches/benchmark_transforms.rs

Skipped

147 µs*

token_classification_postprocess[(64, 512)]

encoderfile-core/benches/postprocessing.rs

Skipped

61.2 ms*

bench_embedding_l2_normalization[(32, 128, 384)]

encoderfile-core/benches/benchmark_transforms.rs

Skipped

65.3 ms*

embedding_postprocess[(8, 16, 384)]

encoderfile-core/benches/postprocessing.rs

Skipped

4.6 ms*

bench_embedding_l2_normalization[(16, 16, 16)]

encoderfile-core/benches/benchmark_transforms.rs

Skipped

337 µs*

Commits

Click on a commit to change the comparison range

Base

main

89ae444

-24.31%

update

25d3e05

3 months ago

by besaleli

+0.01%

remove cargo cache on ci

0c4df94

3 months ago

by besaleli

-0.01%

update benchmark from instrumentation to simulation

ce7dd3e

3 months ago

by besaleli

Home Terms Privacy Docs