blakeslee-research/nucleic-purpose-classifier
3
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

updating purpose-classifier data:

Browse Source
This commit is contained in:
abkslm
2026-08-02 20:15:13 -07:00
parent 98311aa932
commit 0f639bfa05
21 changed files with 14248 additions and 14463 deletions
Download Patch File Download Diff File Expand all files Collapse all files
data/dataset-v1-manifest.json
+76 -79
View File
@@ -1,21 +1,18 @@
{
"curation": {
"excludedDuplicates": {
"near": 18,
"semantic-reviewed": 3
"near": 18
},
"inputRecords": 12214,
"inputRecords": 12007,
"nearDuplicateMethod": "word-trigram Jaccard after SimHash LSH candidate search",
"nearDuplicateThreshold": 0.92,
"retainedRecords": 12193,
"reviewPath": "ml/purpose-classifier/data/curation-review-v1.json",
"reviewSha256": "1721ab77a0f723a4b03f42b14850b5c351ead2cd8fa72bb2d1fe3f0d2bb568da",
"retainedRecords": 11989,
"vagueEvalPolicy": "validation/test only"
},
"datasetVersion": "purpose-dataset-v1",
"datasetVersion": "purpose-dataset-sol-high-v2",
"frozenEval": {
"classifiableShippedFixtureRecords": 87,
"excludedGeneralFixtureRecords": 5,
"classifiableShippedFixtureRecords": 89,
"excludedGeneralFixtureRecords": 3,
"hardSliceDefinition": [
"boundary",
"mixed",
@@ -24,9 +21,9 @@
],
"shippedFixtureRecords": 92,
"shippedFixturesPath": "Tests/NucleicCoreTests/Fixtures/purpose-prompts.json",
"shippedFixturesSha256": "6c8f0eedce35f9da7a407934009e87737b5e85c7371f6a118982d9fae2ea7aae",
"shippedFixturesSha256": "876068ea26d7109365bcfbd1dc44ba3f1a0400c80dc7fdacf008b3fa07cd92cc",
"syntheticPath": "ml/purpose-classifier/data/frozen-test-v1.jsonl",
"syntheticSha256": "5d34c8ac3d39c5b4e2e739c4c9058093beb6f790a42a1397cedefae6271654ab"
"syntheticSha256": "9997dbaa6d305ea56d8c161e39b925368761ab613a706eb8c2ba55d0e58b1906"
},
"ratios": {
"test": 0.1,
@@ -38,113 +35,113 @@
"sources": [
{
"path": "ml/purpose-classifier/data/purpose-prompts.jsonl",
"records": 9214,
"sha256": "3c7f8b496ef2014794a50433a78e3125c0dd4a03dc2a7773b52ee25587ac48e7"
"records": 9007,
"sha256": "6fbd08df113ca4c6f9da151222772dc3e25e1b9b19c9c2b7df9f572053af2507"
},
{
"path": "ml/purpose-classifier/data/purpose-prompts-round2.jsonl",
"records": 3000,
"sha256": "0a35eac2549e95b31518cd0eee99b83b03761c26a0fb258f427fe04a632699ed"
"sha256": "a3c5bbede5b7db743c9f31756491990b388edd2aa9b41c8b2b9ec95b7115f4aa"
}
],
"splits": {
"test": {
"classifiableFixtureRecords": 87,
"classifiableFixtureRecords": 89,
"distribution": {
"language": {
"de": 7,
"en": 1110,
"es": 9,
"fr": 5,
"ja": 3,
"pt": 4,
"de": 6,
"en": 1084,
"es": 7,
"fr": 7,
"ja": 5,
"pt": 7,
"zh": 3
},
"purpose": {
"backendImpl": 171,
"debugging": 149,
"frontendImpl": 157,
"planning": 132,
"quickFix": 140,
"refactor": 136,
"review": 127,
"writing": 129
"backendImpl": 140,
"debugging": 168,
"frontendImpl": 159,
"planning": 120,
"quickFix": 136,
"refactor": 149,
"review": 131,
"writing": 116
},
"slice": {
"boundary": 179,
"core": 507,
"mixed": 80,
"pasted-context": 84,
"vague-eval": 291
"boundary": 51,
"core": 635,
"mixed": 87,
"pasted-context": 77,
"vague-eval": 269
}
},
"hardSyntheticRecords": 634,
"logicalRecords": 1228,
"sha256": "5d34c8ac3d39c5b4e2e739c4c9058093beb6f790a42a1397cedefae6271654ab",
"syntheticRecords": 1141
"hardSyntheticRecords": 484,
"logicalRecords": 1208,
"sha256": "9997dbaa6d305ea56d8c161e39b925368761ab613a706eb8c2ba55d0e58b1906",
"syntheticRecords": 1119
},
"train": {
"distribution": {
"language": {
"de": 100,
"en": 9324,
"es": 100,
"fr": 81,
"ja": 77,
"pt": 70,
"zh": 72
"de": 84,
"en": 9257,
"es": 82,
"fr": 64,
"ja": 62,
"pt": 56,
"zh": 57
},
"purpose": {
"backendImpl": 1226,
"debugging": 1245,
"frontendImpl": 1234,
"planning": 1223,
"quickFix": 1230,
"refactor": 1223,
"review": 1230,
"writing": 1213
"backendImpl": 1223,
"debugging": 1355,
"frontendImpl": 1092,
"planning": 1208,
"quickFix": 1184,
"refactor": 1204,
"review": 1203,
"writing": 1193
},
"slice": {
"boundary": 2066,
"core": 5818,
"mixed": 906,
"pasted-context": 1034
"boundary": 605,
"core": 7175,
"mixed": 987,
"pasted-context": 895
}
},
"records": 9824,
"sha256": "40a5064c2a95d4394d71b3edbad7d03dce7d67b2e5ba9ab9320476e35e3a4953"
"records": 9662,
"sha256": "0e16940308dc7557c73b1b804bf7b715c86d263b613201588ec274169ee01e89"
},
"validation": {
"distribution": {
"language": {
"de": 8,
"en": 1178,
"es": 12,
"de": 10,
"en": 1165,
"es": 10,
"fr": 7,
"ja": 8,
"pt": 8,
"ja": 5,
"pt": 4,
"zh": 7
},
"purpose": {
"backendImpl": 183,
"debugging": 157,
"frontendImpl": 168,
"planning": 141,
"quickFix": 151,
"refactor": 149,
"review": 140,
"writing": 139
"backendImpl": 152,
"debugging": 182,
"frontendImpl": 174,
"planning": 128,
"quickFix": 140,
"refactor": 163,
"review": 143,
"writing": 126
},
"slice": {
"boundary": 195,
"core": 541,
"mixed": 83,
"pasted-context": 95,
"vague-eval": 314
"boundary": 57,
"core": 677,
"mixed": 95,
"pasted-context": 88,
"vague-eval": 291
}
},
"records": 1228,
"sha256": "3d0a3730931db360dd8346cf12d03901ac139acf5be1d9e84975ac591186c588"
"records": 1208,
"sha256": "301cd3d1c69e1bb92b811e857093ad12fb8c5c11bb402ee2993403d5e4467969"
}
}
}