| test bpe-gpt2/encode ... bench: 1,841,401,327 ns/iter (+/- 7,647,747) |
| test bpe-gpt2/encode-batch ... bench: 856,945,817 ns/iter (+/- 2,097,501) |
| test bpe-gpt2/encode-no-cache ... bench: 2,434,969,403 ns/iter (+/- 12,750,446) |
| test bpe-gpt2/encode-batch-no-cache ... bench: 1,005,196,498 ns/iter (+/- 3,388,934) |
|
|
| test llama3/encode ... bench: 1,777,787,252 ns/iter (+/- 8,709,711) |
| test llama3/encode-batch ... bench: 845,861,692 ns/iter (+/- 3,085,809) |
| test llama3/encode-fast ... bench: 1,668,399,123 ns/iter (+/- 11,113,797) |
| test llama3/encode-char-offsets ... bench: 1,058,717,013 ns/iter (+/- 10,209,335) |
| test llama3/concurrent-4t ... bench: 48,088,610 ns/iter (+/- 531,780) |
|
|
| test serialization/from-file-roberta ... bench: 58,454,529 ns/iter (+/- 95,885) |
| test serialization/from-file-llama3 ... bench: 352,452,320 ns/iter (+/- 4,258,030) |
| test serialization/from-file-albert ... bench: 34,679,233 ns/iter (+/- 74,600) |
| test serialization/bpe-from-file-gpt2 ... bench: 42,733,137 ns/iter (+/- 168,038) |
| test serialization/save-llama3 ... bench: 90,189,058 ns/iter (+/- 802,031) |
| test serialization/deserialize-roberta ... bench: 65,634,594 ns/iter (+/- 285,169) |
| test serialization/deserialize-llama3 ... bench: 382,892,586 ns/iter (+/- 4,745,572) |
|
|
| test train/bpe-small ... bench: 16,799,826 ns/iter (+/- 289,242) |
|
|
|
|
