| test bpe-gpt2/encode ... bench: 1,600,956,518 ns/iter (+/- 15,863,420) |
| test bpe-gpt2/encode-batch ... bench: 978,441,503 ns/iter (+/- 2,406,659) |
| test bpe-gpt2/encode-no-cache ... bench: 2,238,737,871 ns/iter (+/- 18,043,470) |
| test bpe-gpt2/encode-batch-no-cache ... bench: 1,128,396,280 ns/iter (+/- 9,500,637) |
|
|
| test llama3/encode ... bench: 1,591,068,009 ns/iter (+/- 19,197,383) |
| test llama3/encode-batch ... bench: 960,859,733 ns/iter (+/- 2,266,412) |
| test llama3/encode-fast ... bench: 1,525,481,196 ns/iter (+/- 7,731,850) |
| test llama3/encode-char-offsets ... bench: 1,151,232,298 ns/iter (+/- 10,322,372) |
| test llama3/concurrent-4t ... bench: 58,487,945 ns/iter (+/- 586,324) |
|
|
| test serialization/from-file-roberta ... bench: 57,770,772 ns/iter (+/- 1,022,131) |
| test serialization/from-file-llama3 ... bench: 380,962,089 ns/iter (+/- 4,213,471) |
| test serialization/from-file-albert ... bench: 37,900,813 ns/iter (+/- 776,479) |
| test serialization/bpe-from-file-gpt2 ... bench: 41,097,612 ns/iter (+/- 388,631) |
| test serialization/save-llama3 ... bench: 90,543,851 ns/iter (+/- 3,672,598) |
| test serialization/deserialize-roberta ... bench: 73,284,461 ns/iter (+/- 1,438,213) |
| test serialization/deserialize-llama3 ... bench: 401,166,081 ns/iter (+/- 5,934,815) |
|
|
| test train/bpe-small ... bench: 14,537,021 ns/iter (+/- 257,484) |
|
|
|
|
