| test bpe-gpt2/encode ... bench: 1,750,547,218 ns/iter (+/- 12,675,592) |
| test bpe-gpt2/encode-batch ... bench: 870,444,459 ns/iter (+/- 4,930,051) |
| test bpe-gpt2/encode-no-cache ... bench: 2,328,584,008 ns/iter (+/- 15,286,190) |
| test bpe-gpt2/encode-batch-no-cache ... bench: 1,006,816,951 ns/iter (+/- 14,723,189) |
|
|
| test llama3/encode ... bench: 1,722,942,840 ns/iter (+/- 12,418,394) |
| test llama3/encode-batch ... bench: 837,015,667 ns/iter (+/- 11,275,382) |
| test llama3/encode-fast ... bench: 1,656,865,835 ns/iter (+/- 25,753,069) |
| test llama3/encode-char-offsets ... bench: 1,057,505,560 ns/iter (+/- 7,934,552) |
| test llama3/concurrent-4t ... bench: 49,461,368 ns/iter (+/- 909,235) |
|
|
| test serialization/from-file-roberta ... bench: 63,651,922 ns/iter (+/- 2,459,535) |
| test serialization/from-file-llama3 ... bench: 365,238,572 ns/iter (+/- 7,669,872) |
| test serialization/from-file-albert ... bench: 37,578,715 ns/iter (+/- 591,419) |
| test serialization/bpe-from-file-gpt2 ... bench: 44,020,472 ns/iter (+/- 720,832) |
| test serialization/save-llama3 ... bench: 110,096,827 ns/iter (+/- 8,859,039) |
| test serialization/deserialize-roberta ... bench: 76,391,206 ns/iter (+/- 3,296,529) |
| test serialization/deserialize-llama3 ... bench: 367,587,893 ns/iter (+/- 8,872,775) |
|
|
| test train/bpe-small ... bench: 18,358,908 ns/iter (+/- 319,163) |
|
|
|
|
