{ "text": "tokenizer/artifacts/heldout_eval.txt", "limit": 10000, "overall": [ { "name": "tiktoken_gpt2", "total_chars": 30775982, "total_bytes": 42250192, "total_tokens": 20869716, "tokens_per_1k_chars": 678.116980962622, "tokens_per_1k_bytes": 493.95553042693865, "bytes_per_token": 2.0244737398438963, "chars_per_token": 1.4746718163294603, "p50_tokens_per_line": 588, "p95_tokens_per_line": 8239, "p95_tokens_per_1k_bytes_per_line": 976.7055970387389 }, { "name": "tiktoken_cl100k_base", "total_chars": 30775982, "total_bytes": 42250192, "total_tokens": 15598584, "tokens_per_1k_chars": 506.8427710933805, "tokens_per_1k_bytes": 369.19557667335573, "bytes_per_token": 2.7085914977923635, "chars_per_token": 1.972998446525659, "p50_tokens_per_line": 552, "p95_tokens_per_line": 5871, "p95_tokens_per_1k_bytes_per_line": 641.7086633069355 }, { "name": "mgpt2_RegexTokenizer_candidate (tokenizer/artifacts/mgpt2.model)", "total_chars": 30775982, "total_bytes": 42250192, "total_tokens": 9485138, "tokens_per_1k_chars": 308.19936143711027, "tokens_per_1k_bytes": 224.49928748252788, "bytes_per_token": 4.454357121635974, "chars_per_token": 3.244653056181154, "p50_tokens_per_line": 469, "p95_tokens_per_line": 3096, "p95_tokens_per_1k_bytes_per_line": 285.07596067917785 } ], "by_bucket": { "devanagari": [ { "name": "tiktoken_gpt2", "total_chars": 1971486, "total_bytes": 5015290, "total_tokens": 2978527, "tokens_per_1k_chars": 1510.8030186367034, "tokens_per_1k_bytes": 593.8892865616943, "bytes_per_token": 1.6838155235792727, "chars_per_token": 0.6618996571123915, "p50_tokens_per_line": 1715, "p95_tokens_per_line": 9396, "p95_tokens_per_1k_bytes_per_line": 615.3846153846154 }, { "name": "tiktoken_cl100k_base", "total_chars": 1971486, "total_bytes": 5015290, "total_tokens": 1928837, "tokens_per_1k_chars": 978.3670794517435, "tokens_per_1k_bytes": 384.5913197442222, "bytes_per_token": 2.6001626887082736, "chars_per_token": 1.0221112514950719, "p50_tokens_per_line": 1112, "p95_tokens_per_line": 6036, "p95_tokens_per_1k_bytes_per_line": 445.6521739130435 }, { "name": "mgpt2_RegexTokenizer_candidate (tokenizer/artifacts/mgpt2.model)", "total_chars": 1971486, "total_bytes": 5015290, "total_tokens": 1102384, "tokens_per_1k_chars": 559.1640011646037, "tokens_per_1k_bytes": 219.80463741877338, "bytes_per_token": 4.549494549993469, "chars_per_token": 1.7883840839489689, "p50_tokens_per_line": 636, "p95_tokens_per_line": 3472, "p95_tokens_per_1k_bytes_per_line": 261.53846153846155 } ], "kannada": [ { "name": "tiktoken_gpt2", "total_chars": 2158481, "total_bytes": 5770626, "total_tokens": 5633881, "tokens_per_1k_chars": 2610.1137790881644, "tokens_per_1k_bytes": 976.3032641519309, "bytes_per_token": 1.0242719006666985, "chars_per_token": 0.3831250606819704, "p50_tokens_per_line": 3068, "p95_tokens_per_line": 19541, "p95_tokens_per_1k_bytes_per_line": 984.8484848484849 }, { "name": "tiktoken_cl100k_base", "total_chars": 2158481, "total_bytes": 5770626, "total_tokens": 3701089, "tokens_per_1k_chars": 1714.6729575103973, "tokens_per_1k_bytes": 641.3669851416466, "bytes_per_token": 1.5591697470663364, "chars_per_token": 0.5832015928284892, "p50_tokens_per_line": 2016, "p95_tokens_per_line": 12796, "p95_tokens_per_1k_bytes_per_line": 645.6692913385826 }, { "name": "mgpt2_RegexTokenizer_candidate (tokenizer/artifacts/mgpt2.model)", "total_chars": 2158481, "total_bytes": 5770626, "total_tokens": 1250792, "tokens_per_1k_chars": 579.4778828259317, "tokens_per_1k_bytes": 216.75152747726156, "bytes_per_token": 4.613577637209064, "chars_per_token": 1.7256914019277385, "p50_tokens_per_line": 689, "p95_tokens_per_line": 4247, "p95_tokens_per_1k_bytes_per_line": 248.54651162790697 } ], "latin": [ { "name": "tiktoken_gpt2", "total_chars": 23547886, "total_bytes": 23664351, "total_tokens": 6049914, "tokens_per_1k_chars": 256.91962327318896, "tokens_per_1k_bytes": 255.65518361352906, "bytes_per_token": 3.911518576958284, "chars_per_token": 3.892267890089016, "p50_tokens_per_line": 423, "p95_tokens_per_line": 2501, "p95_tokens_per_1k_bytes_per_line": 394.47852760736197 }, { "name": "tiktoken_cl100k_base", "total_chars": 23547886, "total_bytes": 23664351, "total_tokens": 5891120, "tokens_per_1k_chars": 250.17617292694555, "tokens_per_1k_bytes": 248.94492141364873, "bytes_per_token": 4.016952803541602, "chars_per_token": 3.9971832181316964, "p50_tokens_per_line": 413, "p95_tokens_per_line": 2433, "p95_tokens_per_1k_bytes_per_line": 378.1904305978987 }, { "name": "mgpt2_RegexTokenizer_candidate (tokenizer/artifacts/mgpt2.model)", "total_chars": 23547886, "total_bytes": 23664351, "total_tokens": 5415481, "tokens_per_1k_chars": 229.9773746144346, "tokens_per_1k_bytes": 228.84553225228953, "bytes_per_token": 4.369759768338214, "chars_per_token": 4.34825383008453, "p50_tokens_per_line": 402, "p95_tokens_per_line": 2171, "p95_tokens_per_1k_bytes_per_line": 288.46153846153845 } ], "mixed": [ { "name": "tiktoken_gpt2", "total_chars": 3098129, "total_bytes": 7799925, "total_tokens": 6207394, "tokens_per_1k_chars": 2003.5944274754215, "tokens_per_1k_bytes": 795.8273957762415, "bytes_per_token": 1.2565538775209049, "chars_per_token": 0.49910300522248147, "p50_tokens_per_line": 5384, "p95_tokens_per_line": 47503, "p95_tokens_per_1k_bytes_per_line": 980.3065905784939 }, { "name": "tiktoken_cl100k_base", "total_chars": 3098129, "total_bytes": 7799925, "total_tokens": 4077538, "tokens_per_1k_chars": 1316.1291863573144, "tokens_per_1k_bytes": 522.7663086504037, "bytes_per_token": 1.9129006277807834, "chars_per_token": 0.7598038326068328, "p50_tokens_per_line": 3720, "p95_tokens_per_line": 31014, "p95_tokens_per_1k_bytes_per_line": 642.8864288642886 }, { "name": "mgpt2_RegexTokenizer_candidate (tokenizer/artifacts/mgpt2.model)", "total_chars": 3098129, "total_bytes": 7799925, "total_tokens": 1716481, "tokens_per_1k_chars": 554.0379370904182, "tokens_per_1k_bytes": 220.0637826645769, "bytes_per_token": 4.5441371037605425, "chars_per_token": 1.8049305526830766, "p50_tokens_per_line": 1719, "p95_tokens_per_line": 13591, "p95_tokens_per_1k_bytes_per_line": 249.65580541532813 } ] } }