climbmix-code-tokenizer / evaluation.json
jfrick's picture
climbmix-code-v1: 32K BPE fitted on the ClimbMix plus code recipe
2d8a612 verified
Raw History Blame Contribute Delete
1.9 kB
{
"eval_sha256": "71e377e1008755b83876ebab70f509df46974387d270ca218ceb7a7263ade543",
"tokenizers": {
"climbmix-code-v1": {
"tokenizer_sha256": "48aa6fb6d01d04067570637ca8a5ecced608f5df5585f9ce44a12c05c5281ae5",
"sources": {
"climbmix": {
"bytes": 14743727,
"tokens": 3217386,
"roundtrip_failures": 0,
"bytes_per_token": 4.582517298204194
},
"code_curated": {
"bytes": 3682462,
"tokens": 1023424,
"roundtrip_failures": 0,
"bytes_per_token": 3.598178272153086
},
"code_mit": {
"bytes": 1733957,
"tokens": 503413,
"roundtrip_failures": 0,
"bytes_per_token": 3.4444025084771352
}
},
"digits": [
"1",
"2",
"3",
"4",
"5",
"\u0120+",
"\u0120",
"6",
"7",
"8",
"\u0120=",
"\u0120",
"1",
"3",
"0",
"2",
"3"
]
},
"structured-code-v1": {
"tokenizer_sha256": "add5b2c7542e4efded329735825ecdcb94df1432617c3d3e4b7508aa459432f9",
"sources": {
"climbmix": {
"bytes": 14743727,
"tokens": 4022037,
"roundtrip_failures": 0,
"bytes_per_token": 3.6657362923314727
},
"code_curated": {
"bytes": 3682462,
"tokens": 1046393,
"roundtrip_failures": 0,
"bytes_per_token": 3.519195942633408
},
"code_mit": {
"bytes": 1733957,
"tokens": 514537,
"roundtrip_failures": 0,
"bytes_per_token": 3.3699364671539658
}
},
"digits": [
"123",
"45",
"\u0120+",
"\u0120",
"678",
"\u0120=",
"\u0120",
"130",
"23"
]
}
}
}