| test bpe-gpt2/encode ... bench: 1,743,907,411 ns/iter (+/- 15,503,925) |
| test bpe-gpt2/encode-batch ... bench: 849,995,940 ns/iter (+/- 16,782,739) |
| test bpe-gpt2/encode-no-cache ... bench: 2,285,958,264 ns/iter (+/- 27,400,846) |
| test bpe-gpt2/encode-batch-no-cache ... bench: 1,014,202,454 ns/iter (+/- 28,848,263) |
|
|
| test llama3/encode ... bench: 1,762,971,908 ns/iter (+/- 22,388,511) |
| test llama3/encode-batch ... bench: 834,771,746 ns/iter (+/- 13,004,285) |
| test llama3/encode-fast ... bench: 1,668,497,254 ns/iter (+/- 21,679,788) |
| test llama3/encode-char-offsets ... bench: 1,044,152,385 ns/iter (+/- 5,792,934) |
| test llama3/concurrent-4t ... bench: 50,147,367 ns/iter (+/- 601,672) |
|
|
| test serialization/from-file-roberta ... bench: 60,238,500 ns/iter (+/- 1,020,688) |
| test serialization/from-file-llama3 ... bench: 358,194,118 ns/iter (+/- 7,466,066) |
| test serialization/from-file-albert ... bench: 39,366,599 ns/iter (+/- 1,377,277) |
| test serialization/bpe-from-file-gpt2 ... bench: 46,117,217 ns/iter (+/- 907,675) |
| test serialization/save-llama3 ... bench: 106,578,372 ns/iter (+/- 2,017,698) |
| test serialization/deserialize-roberta ... bench: 69,488,831 ns/iter (+/- 3,603,382) |
| test serialization/deserialize-llama3 ... bench: 368,773,322 ns/iter (+/- 5,090,980) |
|
|
| test train/bpe-small ... bench: 17,351,662 ns/iter (+/- 741,114) |
|
|
|
|
