fix(asr): restore computer terminology bias
Rebuild the bundled language model and runtime source filtering around a dedicated computer-term corpus so technical dictation keeps domain coverage.
This commit is contained in:
Binary file not shown.
@@ -1,13 +1,14 @@
|
||||
{
|
||||
"bin_bytes" : 41456,
|
||||
"bin_bytes" : 198494,
|
||||
"bin_file" : "OSGKeyboardCLM.bin",
|
||||
"export_seconds" : 0.030717015266418457,
|
||||
"generated_at" : "2026-08-14T15:07:23Z",
|
||||
"export_seconds" : 0.035165071487426758,
|
||||
"generated_at" : "2026-07-06T13:16:27Z",
|
||||
"identifier" : "com.osgkeyboard.custom-lm.v1",
|
||||
"locale" : "zh_CN",
|
||||
"phrase_count" : 3040,
|
||||
"phrase_count" : 13329,
|
||||
"sources" : {
|
||||
"ai_tech_seed" : 3040
|
||||
"ai_tech_seed" : 3040,
|
||||
"computer_terms" : 10300
|
||||
},
|
||||
"version" : "1.0.1"
|
||||
"version" : "1.0.0"
|
||||
}
|
||||
@@ -1,22 +1,22 @@
|
||||
{
|
||||
"entry_count" : 3040,
|
||||
"files" : {
|
||||
"phrases" : "phrases.tsv"
|
||||
},
|
||||
"generated_at" : "2026-08-14T15:07:23Z",
|
||||
"locale" : "zh-Hans",
|
||||
"notes" : [
|
||||
"Project-curated bilingual AI brands, technology terms, companies, and names.",
|
||||
"No third-party cell dictionaries or Sogou-derived data.",
|
||||
"PhraseCount weights map to SFCustomLanguageModelData relative frequencies."
|
||||
],
|
||||
"sources" : [
|
||||
"version": "v1",
|
||||
"generated_at": "2026-07-05T11:27:48.823095+00:00",
|
||||
"locale": "zh-Hans",
|
||||
"entry_count": 10300,
|
||||
"sources": [
|
||||
{
|
||||
"key" : "ai_tech_seed",
|
||||
"label" : "OSGKeyboard curated AI\/tech lexicon",
|
||||
"license" : "MIT (curated seed; OSGKeyboard contributors)",
|
||||
"raw_count" : 3040
|
||||
"key": "computer_terms",
|
||||
"label": "计算机词汇大全【官方推荐】",
|
||||
"weight": 5,
|
||||
"raw_count": 10300
|
||||
}
|
||||
],
|
||||
"version" : "v1"
|
||||
}
|
||||
"notes": [
|
||||
"Domain-specific computer/IT vocabulary only; casual network slang and Sogou popular words removed.",
|
||||
"PhraseCount weights map to SFCustomLanguageModelData relative frequencies.",
|
||||
"Merged with ai-tech-brands seed at export time for the final .bin asset."
|
||||
],
|
||||
"files": {
|
||||
"phrases": "phrases.tsv"
|
||||
}
|
||||
}
|
||||
|
||||
File diff suppressed because it is too large
Load Diff
Binary file not shown.
@@ -1,13 +1,14 @@
|
||||
{
|
||||
"bin_bytes" : 41456,
|
||||
"bin_bytes" : 198494,
|
||||
"bin_file" : "OSGKeyboardCLM.bin",
|
||||
"export_seconds" : 0.030717015266418457,
|
||||
"generated_at" : "2026-08-14T15:07:23Z",
|
||||
"export_seconds" : 0.035165071487426758,
|
||||
"generated_at" : "2026-07-06T13:16:27Z",
|
||||
"identifier" : "com.osgkeyboard.custom-lm.v1",
|
||||
"locale" : "zh_CN",
|
||||
"phrase_count" : 3040,
|
||||
"phrase_count" : 13329,
|
||||
"sources" : {
|
||||
"ai_tech_seed" : 3040
|
||||
"ai_tech_seed" : 3040,
|
||||
"computer_terms" : 10300
|
||||
},
|
||||
"version" : "1.0.1"
|
||||
"version" : "1.0.0"
|
||||
}
|
||||
Reference in New Issue
Block a user