{
  "schema_version": "sasr_eval/0.1",
  "generated_at": "2026-06-15T09:10:35",
  "name": "SASR Eval pilot: Horn of Africa Somali ASR reference-free pass",
  "description": "Exploratory characterization of 4 production ASR systems on Djiboutian Somali audio sampled from 2 publicly broadcast YouTube videos. Reference-free metrics only, ground-truth transcripts have not yet been collected, so no WER/CER.",
  "ground_truth_status": "absent, references pending native-speaker correction of ElevenLabs draft transcripts.",
  "sampling": {
    "clip_length_seconds": 60,
    "edge_buffer_seconds": 60,
    "inter_clip_buffer_seconds": 30,
    "seed_scheme": "seed = 42 + clip_index",
    "note": "BBC clip01 is 107s (user-selected pilot); all other clips are 60s random samples."
  },
  "vendors": {
    "elevenlabs": {
      "model_id": "scribe_v1",
      "lang_hint": "som"
    },
    "whisper": {
      "model_id": "large-v3 (faster-whisper, local)",
      "lang_hint": "so"
    },
    "assemblyai": {
      "model_id": "universal-2",
      "lang_hint": "so"
    },
    "deepgram": {
      "model_id": "nova-2",
      "lang_hint": "multi (so rejected with 400)"
    }
  },
  "videos": {
    "bbc_news": {
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "video_title": "BBC Somali news segment",
      "video_duration_seconds": 830
    },
    "salaf_lecture": {
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "video_title": "Salafmedia, Miisaanka lagu garto Ninka Ragga ah",
      "video_duration_seconds": 1550
    }
  },
  "classification_rules": {
    "empty": "no text returned",
    "cross-lingual misroute": "arabic char share >= 0.50 (output is in wrong language family)",
    "low diversity": "unique-word ratio < 0.45 (severe lexical collapse)",
    "trailing loop": "trailing 12-token n-gram repeat >= 3 OR max consecutive token repeats >= 4",
    "surface-usable": "passes all of the above, HEURISTIC ONLY, not validated against ground truth",
    "not_run": "no transcript exists for this (clip, vendor) pair"
  },
  "limitations": [
    "No ground truth, surface-usable classifications are heuristic. Wait for native-speaker references for any accuracy claim.",
    "N=11 clips drawn from 2 videos = 2 speakers x timestamps, not 11 independent samples.",
    "Single run per (clip, vendor), failure modes are point estimates; determinism unverified.",
    "Deepgram ran on 3 clips only; remaining 8 were skipped after batch hung. All 3 completed runs returned empty.",
    "Vendor language-hint codes differ (som vs so vs multi) because not all vendors accept the same code."
  ],
  "rows": [
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip01",
      "start_seconds": 347,
      "end_seconds": 454,
      "duration_seconds": 107,
      "audio_path": "audio/bbc_news/clip01.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "-dawladduba ay markasta wadahadalka ay wax ayeeyaan ay gardaadinaayeen. Ciidan ciideedii ayaynu ka warqabnaa in Madaxwaynuhu u ballan qaaday in la wadahado intaa soo jeerba la wadahadlay maah oo iyagu keliya ah, waxa weyan muqaarad kamarka laga yimaado imika hubka qaatay. Waxaa kaloo jira muqaarad ku jira in xisaabta iyo iminka doorashada dhababa ka qaybgelaya oo muqaarada oo aan dowladda aan la socon in way badan iyo waliba muqaarad kaasi ah xisbi ah samaysteyn. Waxaa wax weyan magaaladii bayaan ku dhex jiraan dadkii bayaan ku dhex jiraan. Waa dad kala duwan. Markaa horta Madaxweynuhu inta badan uu la faraystay Muqaaradka Imika Hubka qaatay kuwa laba uu la faraystay intaa soo jeer noow la faraystay. Waxay doonayaan inuu waa waadhigaystay. Madaxweynuhu aragti bulshada siyaasadeed oo ku saabsan Dastuurka. Hadday tahay wax weyan in ballanqaadkiisu filiyo ah in la dhameystiro ama wax weyan qabyatir lagu sameeyo Dastuurka wixii ka harsanaa. Iyo haddii ay tahay aragti doorasho oo la doonayo umadan in lagu soo celiyo aragtideeda aragtideeda umadda. Wuxuun bay markaasi fulin kartaa markaa gacanta lagu soo celiyo doorashada. Qof kay doonayaan ay doortaan. Qof kay doonayaan ay diidaan taas oo markaasi Madaxweynuhu uu kaamba qaadayo ayaa waxa lasamaysay gudiyo farabadan gudiyada. Maaha gudiyo dawladu ay leedahay wa gudiyo la doorto. Barlamaankuna iska dhex doortay. Waxa weyn dawladuna ay qayb kula haydo. Waxa weyn gudiyada si ay noqdaan gudiyo la doonayayna dastuurkina ay wax weyn ay wax ka qabtaan doorashada inay wax ka qabtaan runtii gudiyada farax badan leh in badan bay so shaqeeyan. Barlamaanki ba la hor dhigo. Barlamaankuna waa waxa weyan waa golaha keliya ee dadka ay igmadayn. Iyaga afkooda ku hadla caraf. Afcooda ku hadla caraf.",
      "n_segments": 276,
      "words": 276,
      "chars": 1755,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.5761,
      "trailing_3g": 2,
      "max_consec_rep": 2,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip01",
      "start_seconds": 347,
      "end_seconds": 454,
      "duration_seconds": 107,
      "audio_path": "audio/bbc_news/clip01.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "d'uuluddu ba ay mar kas taba wadahadal ka ay uxuweyaan ay qirdaadina iyn. Qeedan, qeeddeedii ayay noo kawar qabnaa in madaxuwinuhu u balan qaaday in la wadahadlo. Intaa soo charba la wadahadlay. Maa hau iyo go keliya uxuweyaan muaarad ka mar ka lagay imaado ima ka hubka qaati waxay ka loo jiraa muaarad ku jiraa iyan ahzabta. U ima ka d'uuruxadibay daba ba ka qibgaliya oo muaarada oo ayn d'uuludda aan lasu'un. Ui badeen u liba muaarad ka si ahzab, hizbiya samaystayn. Waxa waxuweyaan magaladibay ku d'axiraan, d'ad kibay ku d'axiraan waa d'ad kala duwan. Mar ka horta, madaxuwinuhu inta ba u la faraystay, muaarad ka ima ka hubka qaatey ku kalaba. U la faraystay, intaa soo charna u la faraystay, waxay dooneyaan na waa waa dhigaystay. Madaxuwinuhu arak tibuulayim siyaasadayad oo ku saabsan dastuurka. Haday tahay waxuweyaan in balan qaadkis u filiyo ah in la dhammaystaro ah maa waxuweyaan qabiyatir luqu samayyo dastuurka waxay kaharsana. Ya haday ay tahay araktee doora soo oo la dooneyo umaddan in luqu soo u liyo araktee deyda. Araktee deyda umadda waxuun bay mar kaasi fulin karta mar ka qaantan luqu soo u liyo doora sada. Qofka i dooneyaan na ay doortaan, qofka i dooneyaan na ay diidaan. Taas oo mar kaasi madaxuwinuhu u kaanba qaadayyo. Aya waxa la samayyo gu diyo fara badan. Gu diyo di maa ahad, gu diyo duwledu ay la ahay waa gu diyo la doorto, barlamaanku nays ka daxa doortay ah waxuweyaan duwledu na ay qayb ku la ahaydo waxuweyaan gu diyada si aay nukadaan gu diyo la dooney ene dastuurkina ay waxuweyaan ay waxka qabtaan. Daarashad ina ay waxka qabtaan. Runti gu diyada faraha badan le. Inbadan bay soo shaqaynayayn barlamaankiba la hor dhigaay barlamaanku na waa waxuweyaan waa golaa hakiliya ee dadka ay eegi madayn eegi afkuu dhuku ahal dhikara.",
      "n_segments": 36,
      "words": 303,
      "chars": 1770,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.5281,
      "trailing_3g": 1,
      "max_consec_rep": 2,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip01",
      "start_seconds": 347,
      "end_seconds": 454,
      "duration_seconds": 107,
      "audio_path": "audio/bbc_news/clip01.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "نحن نحن نحن نحن نحن نحن نحن نحن نحن نتحدث عن معارض حزبية. نحن نتحدث عن معارض حزبية. نحن نتحدث عن معارض حزبية. مدى حيوانهو عرقتى بولى يمى سياسة دايد اوكو سابسان دستورك هداى تهاى وحويا ان بلنقات كيسو فلو اه ان لما يستروا اما وحويا قبية تير لقوسم ايوه دستورك وحى كهرسنا هداى تهاى عرقتى دو رشو او لدون ايوه امدن ان لقوسو عليو عرقتى دايدة عرقتى دايدة امدن وحومبى مركزى فلين كرتا مركزا قعدن لقوسو عليو دو رشدا وقفقي دونيان اي دورتان وقفقي دونيان مرحباً. أهلاً ومرحباً. أهلاً ومرحباً. أهلاً ومرحباً. نعم ، نعم ، نعم ، نعم.",
      "n_segments": 100,
      "words": 100,
      "chars": 515,
      "script": {
        "latin": 0.0,
        "arabic": 1.0,
        "other": 0.0
      },
      "uwr": 0.567,
      "trailing_3g": 2,
      "max_consec_rep": 9,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip01",
      "start_seconds": 347,
      "end_seconds": 454,
      "duration_seconds": 107,
      "audio_path": "audio/bbc_news/clip01.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": true,
      "transcript_text": "",
      "n_segments": 0,
      "words": 0,
      "chars": 0,
      "script": {
        "latin": 0.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.0,
      "trailing_3g": 0,
      "max_consec_rep": 0,
      "classification": "empty",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip02",
      "start_seconds": 74,
      "end_seconds": 134,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip02.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "Waxaa nabada in isla markaasina ay cabiraan aragtida ay dooneyaan dadka inay ka iibiyaan. Laakiin ah nasib darro waxaa timiruntii in hub la lasoo baxay. Iniman afka khidhay ay hub waawayno BTN iyo kuwi ah lagu qarxin jiray jiibabka iyo tignigkada ad arkayso iyagoo dabada ula soo baxay. Markaa taasi waxay keentay dawladdu markaa iyadoo labadii sano 2020 kii madaxweynuhu markii uu kursiga uu u qabtay Hogaanka Waxay u ka go'noonayd inuu nabad ku soo waaro hambaba ku soo dabaalo dhammaan Somalia. Taasi waxay keenta markaa afartii sanno ee u xiiso wadahad ah, bashbash ka ahayd in la yiraahdo maalin keliya lagu ciyaaraya dib loogu celiyay. Wuxuu ugu yaraa in aan ka dhexgelo ee madaxda ama mucaaradka si qaar kood ay hayeen Madax xurruudanka Soomaaliyeed. Waxay gun ku cedeeyeen dowladdu inay iyo ay soo weerartay in.",
      "n_segments": 136,
      "words": 136,
      "chars": 819,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.7353,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip02",
      "start_seconds": 74,
      "end_seconds": 134,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip02.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "islamarkaas inaay aabiraan arak ti doda ay dooneean dadka ina kay ibiyaan. Lakiin nasibdarro waxaa timirhunti in hub la lasoobaxay. In man afka xaday ay hub waa weeno BTN iyo kuwi loogu waxa weyaan na qarxan cheeray in chiibibka iyo ticnikaada. Aad aarkay sooyego dabado la lasoobaxay markaa. Taa ziwaxay kaante doolado markaa iyo dode inbadaan ilabadi kuniya labeela qarxan ki midaxuun u marki u kursiga u waxa weyaan u qabtay hugaanka. Waxaay u kaggoonayd unu nabad ku soo waro, ambabaw ku soo dabaalo laman Somaliya. Taa ziwaxay kaante markaa afar tiisano. Waxiil le soo wadi, e nabad ahayde, e bashbashka ahayd. Ila ilda hado maaliin kariyaa loogu u iyaariyaay. Wataay lakiin Aad Arman, hadii aankar naaxkalo madaxada amaa muaaraadkaasi waa qaarkood ay haayan madax waraddaar ka soo maamulay. Waxaay eegu nukuaade induguludu in iyaadu ay soo waratee.",
      "n_segments": 17,
      "words": 138,
      "chars": 854,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.7391,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip02",
      "start_seconds": 74,
      "end_seconds": 134,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip02.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "ونحن نتعامل مع أهل المنطقة الإسلامية لتعامل مع أهل العالم. لكن نصيب درّو وحاة تمرنتي إن هب لأصوّب حي إن من أفك حدّي أي هب ووينو BTN يوكوي لغوّو وحويا نقرحنجريين تشيببك ايوووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووووو ونحن نتحدث عن نبات وكيفية تصويرها وكيفية تصويرها في المنطقة الصومالية. ونحن نتحدث عن نبات وكيفية تصويرها في المنطقة الصومالية. نتحدث عن ونحن نبات وكيفية تصويرها في المنطقة الصومالية. ونحن نتحدث عن نبات وكيفية تصويرها في المنطقة الصومالية. ونحن نتحدث عن نبات وكيفية تصويرها في المنطقة الصومالية. ونحن نتحدث عن نبات وكيفية تصويرها في المنطقة الصومالية. ونحن نتحدث عن نبات وكيفية تصويرها في المنطقة الصوم",
      "n_segments": 98,
      "words": 98,
      "chars": 718,
      "script": {
        "latin": 0.0049,
        "arabic": 0.9951,
        "other": 0.0
      },
      "uwr": 0.3922,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip02",
      "start_seconds": 74,
      "end_seconds": 134,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip02.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": true,
      "transcript_text": "",
      "n_segments": 0,
      "words": 0,
      "chars": 0,
      "script": {
        "latin": 0.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.0,
      "trailing_3g": 0,
      "max_consec_rep": 0,
      "classification": "empty",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip03",
      "start_seconds": 198,
      "end_seconds": 258,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip03.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "In ammaanka dalka ay wax weeyan loo dheeriyo inay ammaanka dalka ilaaliso ama gobolka ilaaliso. Ammaanka Muqdisho ulba sigaar ahaneed oo caasimadiya. Caasimadu wax weeyan waa ammaan markaa caasimada ammaan noqotaana waa somaliga ammaan waa Somaliyo qurux badan. Markayna caasimadii hubnu caasa lala soo baxo taasi iyaga belaabay iyagaana belaabay in hub waaweyn oo aan markii horeba aan loo ogolayn in lala soo baxo xeer baa la sameymo saaxiibin doogteynta kale. Xeer kaasi wax lagaga xaakamaynta hubka iyo runtii hore ugu celiyay. Dowladduna waa wad an saxeysay. Xeer kaasi waa xeer lagu talagalay in dhammaan hubka waaweyn oo dhan aan magaalada caasimada iyo meelaha kala ba aan lala marin. Ninkastoo madaxweynuhu kaantoroolo hubka ula marayo oo xubno hub waxa weeyan aad u fudud mar hadii amaan laga dhigay Muqdisho. Waa yahay laakiin waxa ay iga gudahay ku eedeeyayaan in ciidamadii qaranka iyo tangiyada dabaabka.",
      "n_segments": 144,
      "words": 144,
      "chars": 918,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.6875,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip03",
      "start_seconds": 198,
      "end_seconds": 258,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip03.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "An in Ammaanka dalka waxuweyaan loo daarii, in Ammaanka dalka ilaaliso, Ammaanka qobalka ilaaliso, Ammaanka Mugdishu ulba sigaarahaanayad oo aasimaddiya, aasimaddu waxuweyaan waa Ammaanka, maraka aasimaddu Ammaanka qotana waa Somaliyo Ammaanka waa Somaliyo quroxbadaan. Maka aasimaddii hubnu aasa lalasoo baxo taas iyaaga bilaabay, iyaaga aana bilaabay in hub waa ween oo aan markeeyo ahra baan loo gulayn in lalasoo baxo xeer baa la samimuayi, saalimeeyna dhuktiyeen kale, xeer kaasi waa xeer dhaa xakamaynta xubka. In xrunteeyna gulaha wu siiradaa xora oo aansihiyay, duuluddu na waa wada aansihisay, xeer kaasi waa xeer lugu tala gulay, nindab maan xubka waa ween oodan, aan magaalada aasimaddaa iyaa meelaha kalaba aan la lamariin. Ning kastoo maada xeenaa uu kaantaroo loo xubka uu lamarayoo oo uu noogda xub waa xa ween aadu futad maraaddii aman lagaadigaay Muqtishoo. Waa yahaay, lakin waa yago noo idhuku aadhayna yaan in aadhamadhi qaranka ya tangiidi id babadki.",
      "n_segments": 5,
      "words": 148,
      "chars": 972,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.6689,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip03",
      "start_seconds": 198,
      "end_seconds": 258,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip03.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "ونحن نتحدث عن أمان. نحن نتحدث عن أمان. نحن نتحدث عن أمان. نحن نتحدث عن أمان. ونحن نتمنى أن يتعاملوا مع بعضهم البعض ونحن نتمنى أن يتعاملوا مع بعضهم البعض نحن نحن نحن نحن نحن نحن نحن نحن",
      "n_segments": 38,
      "words": 38,
      "chars": 184,
      "script": {
        "latin": 0.0,
        "arabic": 1.0,
        "other": 0.0
      },
      "uwr": 0.2895,
      "trailing_3g": 6,
      "max_consec_rep": 8,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip03",
      "start_seconds": 198,
      "end_seconds": 258,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip03.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": false,
      "transcript_text": null,
      "n_segments": null,
      "words": null,
      "chars": null,
      "script": null,
      "uwr": null,
      "trailing_3g": null,
      "max_consec_rep": null,
      "classification": "not_run",
      "notes": "Deepgram batch terminated after 3 hung calls; failure mode established on completed clips."
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip04",
      "start_seconds": 513,
      "end_seconds": 573,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip04.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "In wax weeyan meeshu aanay ahayn bilowday orgiga ka wayn ee bilayada ya bilaabay, iyo bilayada bilaabay muqaadka bilaabay marka ay hub la soo baxeen marka maalmihii lagu talo galay ahayd ay ka mudaharaadan ay igu diideen meelo lagu talo galay baajir ah oo la doonayo ammaanka Mogadishu inay ku sii adkaato. Isla markaasana uma doonayo arkaan hadii ba ay markaasi ay isla hayaan shacabka aragtidisa ay markaasi matali karaan. Tiina waan markaas wayn soroo buu day. Waa gartay Turkigu ayaa lala yahay safka hor umay kuuga jira arrimaha dhexdhexaadinta. Waa maxay kaalinta Turkiga arrimahan hadda taagan sidii xallu gaarilaha. Runtii dawlada Turkigu waa dowlad aynu saaxiibna hay. Waa dowlad usoo gurmay somalia maalmihii abaaraha isla markaasana sidii u joogta. Wax badana Somalia kala qabatay haday tahay Ammaanka iyo haday tahay Horumarkeeda iyo haday tahay kaaba Yaasha Dhaqaalahaba. Marka hadii ba, waxa weeyan markasta dawlada Turkigu iyo dawlada Somaliya ugu dashashqeysa.",
      "n_segments": 151,
      "words": 151,
      "chars": 976,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.7086,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip04",
      "start_seconds": 513,
      "end_seconds": 573,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip04.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "in waxooyan meyshu aanay ahayn barortu urgiga kawin ee bilaayada yaa bilaabay yaa bilaayada bilaabay muaaradka bilaabay marka ay hubla soo bixayn marka meylee hii loogu tala qalaya ahaydayn ka mudaahraadan iyo guddiydeyn meylea loogu tala qalaya bachira o la doonayi amman ka mugdishana inay ku si aadkaato isla markaasana umad doonayi arkaan hadiba ay markaasi ay isla hayyan shaabka arakdidiysa ay markaasi matali karan tiina warkaasu u suroobu ude wagaartay turkiga ayaa la layahay safka u raayu ka gajra arimaha d'aad haadinta wama xay kaalinta turkiga arimaha n'hadda taa gan sidi xallu wagaarilaha runti dowlad turku guwa dowlad ayno sahibna hay wad dowlad u soo gurmata Somalia maalmihi abaaraha isla markaasana sidi uchoogta waxbada na Somalia kalakabatay adaytahay amman ka iya adaytahay horumarka iya adaytahay kaabayashada qalahaba markaa hadiba waxawiyan markaas tabad dowlad turku ga iya dowlad Somalia ugu dasha qay",
      "n_segments": 32,
      "words": 142,
      "chars": 929,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.7254,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip04",
      "start_seconds": 513,
      "end_seconds": 573,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip04.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ونحن ونحن نتعامل مع بعضنا البعض ونحن نتعامل مع بعضنا البعض ماذا يجب أن نفعل لتحقيق تركيا؟ دولة ترك ودولة ساحبنا هي ودولة اسو قرمتي صومالية معلمه عبارها اسلام مركزنا سيذيو تشوكتا وحفظانا صومالية كالقبطة عديدة اي امان كاي عديدة اي هرومر كاي عديدة اي كابايا شنقالهبا مركز ترك ودولة صومالية",
      "n_segments": 106,
      "words": 106,
      "chars": 583,
      "script": {
        "latin": 0.0,
        "arabic": 1.0,
        "other": 0.0
      },
      "uwr": 0.3302,
      "trailing_3g": 1,
      "max_consec_rep": 2,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip04",
      "start_seconds": 513,
      "end_seconds": 573,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip04.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": false,
      "transcript_text": null,
      "n_segments": null,
      "words": null,
      "chars": null,
      "script": null,
      "uwr": null,
      "trailing_3g": null,
      "max_consec_rep": null,
      "classification": "not_run",
      "notes": "Deepgram batch terminated after 3 hung calls; failure mode established on completed clips."
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip05",
      "start_seconds": 631,
      "end_seconds": 691,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip05.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "-dalkaaga reebaynayo, cambareenaya, waxa weey madaxweyne ku leh oo soo dhawaynaya markasta oo wada hadalka. Waa gartay dadkii qaar waxay qabaan dalkani waa mid nugul siyaasadahaan. Marka yacni waxaan u baahan yahay in la isku wado oo yacni maaratay inta nuglaashaha si laga baxaayo dadki la isku soo jeedzido halkii gacan bir abhal ugu qaban lahaa. So uma muuqato in qaabkii soo jireenka ahaa ee markaasi siyaasiyan dalku uu ku soo jiray ama markay timaadato xagga doorashada loogu gudbi jiray in adka leexateen. Runtii markastaba dee, dawladdu iyadaa u wadayba in la isku wado. Madaxwaynaha now shanaa aye Xasan Sheikh Mahamud afartiisan u jooga iyo horumba inuu yahay ninka keliya ugu waayee. Lisan isagoo madaxwaynaha ah oo awood badan gacanta ku hayo markastaba la'aha inuu wada doodno inuu wada hadalno. Waxaa la isku hayaa saali minna waa doorasho. Doorashadiina waa one man vote. Dastuurkiiba markaasi cusbaa bay ku qoran tahay Baarlamaanka baa si xigay. Kulehyaha qalalka saraaca si xigay.",
      "n_segments": 159,
      "words": 159,
      "chars": 997,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.761,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip05",
      "start_seconds": 631,
      "end_seconds": 691,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip05.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "qaabaa qayabaynayyo, aambaraynayyo, waxaway ma daxuunoo kulli buu suuduwaynaya markaas dhugu dhaha dhalka. Waa qirtaay dhadki qaar waxay qabaan dhalkan ii waamid nuguul siyasa dhaxaan. Markaa, wixan oo baahay ahay inlays ku wado, oo maa rattaay inta nugu laxay ahasi lagaabaxay dhadki ilays ku soojii jido. Halke qaambiraa waxa lagu qabaan laxaan. So, u moom u qatooyin qaab ki soojiirayn ka ahaa markaasi siyasiayn dhalku u ku soojiiray. Amaa markaay timaado xaaga dhawrashada loogood bichiray inaad kalla xateen. Runti, markaas tabaad ee doolado iya dhaabaa waddeeba inlays ku wado. Madaxuuynaa nawa xanaa misan ahay, xaasan sheeqa Mahamood afartii san u u chooga iya horaba, inu ya ahay ninka kali u guu ahay ilsan isago madaxuuynaa. Oo awood badan qaanta kuhaa iya inu markaas tabaada ahay inu wada doodno, ay inu wada hadalno. Waxalays kuhaa iya Salimoo noo waa dhawrashad. Dhawrashada ina waa wanmaa woon food. Dastoor kiiba markaasi usbabaay ku qoranta ahay. Parlaman kiiba zihyay. Kolohi aqal ka sarayan zihyay.",
      "n_segments": 19,
      "words": 160,
      "chars": 1019,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.7625,
      "trailing_3g": 1,
      "max_consec_rep": 2,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip05",
      "start_seconds": 631,
      "end_seconds": 691,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip05.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "نتمنى لكم جميع حلقاتكم في مجتمعنا. نحن نحن نحن نحن نحن نحن نحن نحن مركز الدولة الدولية مدحوينا نوحنا مصنعي حسن شيخ محمود أفرطي سنوتي وغيرها مدحوينا نوحنا مدحوينا نوحنا مصنعي حسن شيخ محمود أفرطي سنوتي وغيرها مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدحوينا نوحنا مدح",
      "n_segments": 64,
      "words": 64,
      "chars": 392,
      "script": {
        "latin": 0.0,
        "arabic": 1.0,
        "other": 0.0
      },
      "uwr": 0.3125,
      "trailing_3g": 5,
      "max_consec_rep": 8,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "bbc_news",
      "domain": "broadcast_news",
      "clip_id": "clip05",
      "start_seconds": 631,
      "end_seconds": 691,
      "duration_seconds": 60,
      "audio_path": "audio/bbc_news/clip05.wav",
      "video_url": "https://www.youtube.com/watch?v=60IIS9dh0jk",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": false,
      "transcript_text": null,
      "n_segments": null,
      "words": null,
      "chars": null,
      "script": null,
      "uwr": null,
      "trailing_3g": null,
      "max_consec_rep": null,
      "classification": "not_run",
      "notes": "Deepgram batch terminated after 3 hung calls; failure mode established on completed clips."
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip01",
      "start_seconds": 843,
      "end_seconds": 903,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip01.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "Waxay tilmaamaysaa muwaajiintii meesha lagu hayo mizaana hane heyneenu nabiga Muhammad aleysalatu wassalaam. Wax ku qiimaynayo in baraya saxaabadatan bulshadeed iyo faalliyuuskeeda in la beddelo. Raqmadii jaahiliyadda laga leeyahay in muuqaalada iyo wax xadadku haystaan iyo sidde loo eg yo wuxuu ku qiimeynaa inaanu muslimka wuxuu ku qiimeynaa. Waxaad dadka lagu qiimayaa ma kursigaa haddaa, iyo siiaadada iyo mansaq iyo jaahya. Inuu madaxweyne iyo waasiiriya, hay kan iyo hay makaasba. Nabigu caleefka salaatuhu wasalaamu ma miil fadhigan nin soo maray. Barka sunnin fadhigin leh fadhigin, maan ra'yuka fi haada. Ninka halkaa maraya. Xukulayi. Rawaa'da qaarkoodna Ma taqooluna fi haada buydi. Ninkaha maxaa odhan lahaydeen.",
      "n_segments": 104,
      "words": 104,
      "chars": 725,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.7981,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip01",
      "start_seconds": 843,
      "end_seconds": 903,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip01.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "waxay teelmaa misaa mawaazinti meeshal logo hayay miizanaa hana hayn inu nabii Muhammada a.s. waxkuu qiimaynayyo, inu baray ya sahabada taan bulshadu vaaliyuska dayin labbedaloo. Naqamadi jahiliyada lagalayimi, emu qalada iyo waxadad ku haystaan iyo sida loo agi waxkuu qiimaynayna, anu muslimku waxkuu qiimaynayna. Waxadad ka loogu qiimayna iyo ma kursiga ahadda, iyo siyadada iyo mansak, iyo jahya. Inu mada xuwe inay ay waziri ya hay kanya, hay makaasba. Nabiga a.s. wama yalfa dayan nisaa maray. War kasuun nim faddii, lim faddii, dimaa raayuka fii hadaa? Nikahaal ka maray yaa? Hu ku layi? Rawaayda qaar kood naa maa taqooloo naa fii hadaa buiidi. Nikahaa mahaa odaan la haydein.",
      "n_segments": 10,
      "words": 108,
      "chars": 684,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.8,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip01",
      "start_seconds": 843,
      "end_seconds": 903,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip01.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "وحقوق قيمين نبيه عليه الصلاة والسلام وما رأيك في هذا ؟ ماذا تقولون بهذا ؟",
      "n_segments": 15,
      "words": 15,
      "chars": 73,
      "script": {
        "latin": 0.0,
        "arabic": 1.0,
        "other": 0.0
      },
      "uwr": 1.0,
      "trailing_3g": 0,
      "max_consec_rep": 1,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip01",
      "start_seconds": 843,
      "end_seconds": 903,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip01.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": true,
      "transcript_text": "",
      "n_segments": 0,
      "words": 0,
      "chars": 0,
      "script": {
        "latin": 0.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.0,
      "trailing_3g": 0,
      "max_consec_rep": 0,
      "classification": "empty",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip02",
      "start_seconds": 175,
      "end_seconds": 235,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip02.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "In Allah lasta bi kasid. Ama anta indallahigal hadi shama ilaha muhammedya ku yaala hadis sahih shahih. Nabi ku waram ilahi aktiisa mid bari ayaa maamid aan kadma ininka ma tihiin u nabi ku qayb gay wa salam day inaadagum ka gaar kaada kadma in laasta bi ka sidan indallahu iyo nabi ku qayb gala salam idam quruxodu iyo jamalka iyo muqaalkaan sare. Haddii aktiisa waxa qiimaha kule maaha saynukala qurux badana ilaha uguma kale duwin saynukala cadna heygu ma kale duwin ilaha aktiisa wax kale qeema kale. Sidoo kalana xaddidki imamu muslimu warina ayuu nabi ku haysay rubba ashcaafa madfuu'in bil abwab la waqtsama ala allah illaa barra nabi mami alehissalaam.",
      "n_segments": 112,
      "words": 112,
      "chars": 660,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.7589,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip02",
      "start_seconds": 175,
      "end_seconds": 235,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip02.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "in nda Allahi lasta bi kaasid ama anta nda Allahi qaal hadidku shama'i ilal muhammadiyya buku iya allaa qana hadis saheeh, asa'i ulamu dhu sha'aqa'i buku nabu ku ilda alayhi salatu wa salam ilahay aqti samid baariya ya'a mammid aan qadmaynin ka mati hiddu nabu ulayi alayhi salatu wa salam day binay adam ku muqaal ka qawar kaanana aad qadmaynin lasta bi kaasid in nda Allahi ibu nabu ulayi alayhi salatu wa salam idaan quruxdu iya jamaalka, iya muqaal kaan sara'i allaa aqtisa waha qiimaha ku la maaha sayin uka la qurux badanna ilahi ugu makala dawin sayin uka la aadmaha ugu makala dawin ilahi aqtisa waha kala qiimaha kale isda waga dda hadidki imamu muslimi uwarinayi uxu nabu ulayi rubba asha'atha madfu'in bila buwab lawaqsa ma'aal Allahi la barra nabu ku tilmaabu qadmayna qadmayna qadmayna qadmayna",
      "n_segments": 24,
      "words": 137,
      "chars": 807,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.6496,
      "trailing_3g": 2,
      "max_consec_rep": 4,
      "classification": "trailing loop",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip02",
      "start_seconds": 175,
      "end_seconds": 235,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip02.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "لست بكاسد عند الله اذن قرح ذو اي جمالك اي موقع الكنسر اللي اكتسوها قيمها كلماها سينوكا لقرح بضننا الهي وقمكا للدوين سينوكا لعضنا هي وقمكا للدوين الهي اكتسوها كلا قيمكا له شدى وقيدوى حديث كي امامه مسلم ورينيه وحنابه قوليه رب اشعث مدفوع بلا بواب لو اقسم على الله اللي ابره حنابه قلت لمامي عليه السلام",
      "n_segments": 57,
      "words": 57,
      "chars": 298,
      "script": {
        "latin": 0.0,
        "arabic": 1.0,
        "other": 0.0
      },
      "uwr": 0.8596,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip02",
      "start_seconds": 175,
      "end_seconds": 235,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip02.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": false,
      "transcript_text": null,
      "n_segments": null,
      "words": null,
      "chars": null,
      "script": null,
      "uwr": null,
      "trailing_3g": null,
      "max_consec_rep": null,
      "classification": "not_run",
      "notes": "Deepgram batch terminated after 3 hung calls; failure mode established on completed clips."
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip03",
      "start_seconds": 380,
      "end_seconds": 440,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip03.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "Waxaa qalbiga idinkaga jireen rumaysantiin iyo waxay idinkasoo fulaya ee camaala. Waxay idinkasoo fulaya camala noocawiyahay. Iyo waxaa quluubti na ku jiraa kaasbaan rabbi qiimaya muu Nabigu tilmaamaya caleesa la'aatu wasalam. Sida waqti loo xirta qisacaana saxaabaa jiray isna ragga aan quruxda badnayn ka mid ahaa. Waxna hay sannin marka waa nimani quruxna badnayn meel duniyana lahayn. Markaasuu Nabigu ay dayi Juleybiib ba leeyahay, hawar Juleybiibu Xaad u guursan wayday,oya iyo guurinayay buu dhacan. Maallamelihi ya'ni baa ka muuqada in qurux la iga raacan igama muuqato gacanta idina wax kuma jira. Iyo markaa gabadhii guurinaya. Nabigu caleesa la'aatu wasalamu waaxay qola ansaarabu utage.",
      "n_segments": 103,
      "words": 103,
      "chars": 698,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.875,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip03",
      "start_seconds": 380,
      "end_seconds": 440,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip03.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "waha qalbi qaydinka qaciray arru maysantin iyo waha qaydinka soofulayya iyo a'mala waha qaydinka soofulayya a'mala noo'awiyahay iyo waha qulubtina ku carayka asbaar rabbi qiimayaya munnabu qutil maami a.s. sida waha qaydika qacirtaq qisa caana sahaaba caray isna raqaam qruxda badnayin ka midaha waha na haysanin marqa waha nemaan qruxna badnayin maal a dunyana lahayin marqa soo nabu quayde julaybiba layda war julaybibaw xaadu gur samu quayde oo iya iyo gur inayabu udaaniga maal la malihi yaani baaka moogada eyn qruxla iyo gur raa'anayiga waha moogada qaanta iyo dana waha kumacira iya marqa qabadi iyo gur inayaya nabu quay alayhi salatu wassalamu waha nabu quay alayhi salatu wassalamu waha qola ansaarabu utabi",
      "n_segments": 29,
      "words": 112,
      "chars": 717,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.6726,
      "trailing_3g": 2,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip03",
      "start_seconds": 380,
      "end_seconds": 440,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip03.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "وحا قلبي قايدن كاجيريا الرومي سنتين ايه وحا قايدن كسوفولية اعمالة وحا قايدن كسوفولية اعمالة نوعو ايه هاي ايه وحا قلوبتين نكو جيرا كهاسبار ربي قيمة ايه يا ابو نبو قطير المامي عليه السلامة وسلامة ستابو قايدو حاج جيرتاق سعانة سحابة جيري اسنا رقام قروح ضبضنين كم ايضاها وحنا هاي سنين قلت لهم انا لن اتخذ القروح انا لن اتخذ القروح انا لن اتخذ القروح انا لن اتخذ القروح نبوه عليه الصلاة والسلام وبحي قل عن سار بوته",
      "n_segments": 78,
      "words": 78,
      "chars": 408,
      "script": {
        "latin": 0.0,
        "arabic": 1.0,
        "other": 0.0
      },
      "uwr": 0.6923,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip03",
      "start_seconds": 380,
      "end_seconds": 440,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip03.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": false,
      "transcript_text": null,
      "n_segments": null,
      "words": null,
      "chars": null,
      "script": null,
      "uwr": null,
      "trailing_3g": null,
      "max_consec_rep": null,
      "classification": "not_run",
      "notes": "Deepgram batch terminated after 3 hung calls; failure mode established on completed clips."
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip04",
      "start_seconds": 470,
      "end_seconds": 530,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip04.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "Waxaan lehay. Ninka na waxa u yimi Nabigii oo leh inantada isi iyo yi guuri. Dabatana nicim ma waa nuucma caynimu buyri. Yaani waa erayo tilmaamaya si uu ugu faraxsan yahay. Waa khayru bushaaraysanayo. Waa wixii la ku doogsataa buuleeyey. Nabiga Muhammad oo gabarnaga soo doonay Aleysalaatu Wasalaam. Markaasuu Nabigu wuxuu kuu yheeyay inay lastu li nafsi ureyduha. Anigun la fahayga uma doonayabu Nabigu calo salaat wasalaam. Gabartan ku waydiisanayo aniguma aha qofka guursanaya qaala faliman buyuridu ilahi rasoolkiisa hada ayadoo doonaysa ergayga ka tag. Qaalali Juleybiibi Nabigu juleybiiban erga wahayd. Markaaninkii wixii u moodayay waxaan ahayba asoo baxay.",
      "n_segments": 99,
      "words": 99,
      "chars": 665,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.8788,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip04",
      "start_seconds": 470,
      "end_seconds": 530,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip04.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "Ninkan na wahaabaw yimeen nabigiyo leh, inan tada isiyo yi guuri, dabaytana ni'imma wa nu'ma aayni buuli. Yaani waa erayyo tilmaa maya suu ugu fura ahsan yahan, waa de khayru bishari sanayyo, waa wahili kuddox se tabuulayyi. Nabi Muhammad oo qabar naga sodoanay aalaihi salatu wasalam. Marka suu nabigoo uku uri, inni lastu li nafsi uriidu haa. Anugun lafahiga umet doonayabu nabigoo aalaihi salatu wasalam. Qabardan kuwaydi sanayyo, anugun maaa ahaa qofka guur sanayya. Qaala fali manbu uri, ilahi rasul kisi yu haddaa iyaadu doonaysa ergayga kattay. Qaala li julaibibinbu nabigoo uri. Julaibiban erga wahaay. Markaaninki wihi u muudayay, wahaaray. Anaheim ba sooba hai.",
      "n_segments": 11,
      "words": 104,
      "chars": 671,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.8654,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip04",
      "start_seconds": 470,
      "end_seconds": 530,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip04.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "نعم ونعم عيني يعني واريو تلمام اياه شو اغفر حسين واده خير وبشاري سري ووحي لكودوك ستابولي نبي محمد وقبرنقا صلى الله عليه وسلم مركز النبي صلى الله عليه وسلم قال اني لست لنفسي اريدها وانا لفهي ومدوني ابن النبي عليه الصلاة والسلام وانا لفهي ومدوني ابن النبي عليه الصلاة والسلام وانا لفهي ومدوني ابن النبي عليه الصلاة والسلام وانا لفهي ومدوني ابن النبي عليه الصلاة والسلام وانا لفهي ومدوني ابن النبي عليه الصلاة والسلام وانا لفهي ومدوني ابن النبي عليه الصلاة والسلام وانا لفهي ومدوني ابن النبي عليه الصلاة والسلام وانا لفهي ومدوني ابن الن",
      "n_segments": 96,
      "words": 96,
      "chars": 533,
      "script": {
        "latin": 0.0,
        "arabic": 1.0,
        "other": 0.0
      },
      "uwr": 0.3958,
      "trailing_3g": 2,
      "max_consec_rep": 1,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip04",
      "start_seconds": 470,
      "end_seconds": 530,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip04.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": false,
      "transcript_text": null,
      "n_segments": null,
      "words": null,
      "chars": null,
      "script": null,
      "uwr": null,
      "trailing_3g": null,
      "max_consec_rep": null,
      "classification": "not_run",
      "notes": "Deepgram batch terminated after 3 hung calls; failure mode established on completed clips."
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip05",
      "start_seconds": 614,
      "end_seconds": 674,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip05.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "La amrillahi la zawijuu julaybiba ilaahima khudharteed julaybi baan ugu weynatay daasiin maaya bay salaantii celi. Odoygii marada islaanti ka hortimi uu kacay markii sidi xagii nabiga qabaa aleysalaatu wasalaam inoonti way maglayso warka dhex socda aabiheedyahood yadeen ayey xaga damakale hadashay. Waxay ku diriy wa xaan idiin waydiyo yaaga kiin soo doonay markaa saa leydhay rasuulki ilaahya aleysalaatu wasalaamu saytidi attaru duuna ala rasuulillaahi sallallaahu alehi wa aalihi wasallama amra uma nabi muhammad bahay dal ku celinaysaan markaa intaan nabi muhammadiidiin imaa u yidhi gaba dhaysiiya miyaa haddana nabigii celinaysaan bay ku tilly aniga nabi iguu wareyciya bay tilly waaye nabi muhammad mar naba aniga ay dayaci maayo ee hadiidina warka nabi gaal aleysalaatu wasalaamu.",
      "n_segments": 116,
      "words": 116,
      "chars": 789,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.8362,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip05",
      "start_seconds": 614,
      "end_seconds": 674,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip05.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "La la'amru Allahi la'uzawwiju juleybiba ilaahi maqurdaarti juleybiba ngu ina tayda siin maayabay salanti tiri odayi maradayi salanti ko hortimi uuqa'ay marku isidi haqqi nabiga qaba aleyhi salatu wasalaam inaanti waa'i maghlay saa warkar dahi so'o dhaaa bihadeeho yadee ayay haqa dhamakala hadas shay waxay ku tiri wahanidin wadiyo yaa ega kin so'o doonay marka saa laydi rasulki ilaahi aleyhi salatu wasalaam saytiri atarudduna a'la rasulillahi sallallahu aleyhi wa aleyhi wasalaam a'mru oo ma nabii Muhammad ba ahadalku ili nisaan marka te nabii Muhammadideen yimawidee qabar da isiiya miyaa haddane nabigi ili nisaanbay ku tiri aniga nabiga igu urechiyabay tiri waa'i nabii Muhammad mar naba aniga idayi imaayyo iyaa nabii Muhammad mar naba aniga idayi imaayyo iyaa nabii Muhammad mar naba aniga idayi imaayyo iyaa nabii Muhammad mar naba aniga idayi imaayyo",
      "n_segments": 19,
      "words": 132,
      "chars": 861,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.6515,
      "trailing_3g": 2,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip05",
      "start_seconds": 614,
      "end_seconds": 674,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip05.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "لا لعمر الله لأوزوج جليبيبا إلهي من خردارتي جليبيبا وإن أطير سينما يبقى إسرانتي تليه وضعي مرة دا إسرانتي كهرتمي ووقعي مركوسيني حقي النبي قبل عليه الصلاة والسلام إن أنتي وي مغلي ساه وركر نحص عذاء بهديه ويده أيه حقا دا مكله هذا الشيء ونحن نحن نحن نحن نحن نحن نحن نحن نحن نبي محمد مرنبا يضيع مياه",
      "n_segments": 58,
      "words": 58,
      "chars": 293,
      "script": {
        "latin": 0.0,
        "arabic": 1.0,
        "other": 0.0
      },
      "uwr": 0.8276,
      "trailing_3g": 5,
      "max_consec_rep": 8,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip05",
      "start_seconds": 614,
      "end_seconds": 674,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip05.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": false,
      "transcript_text": null,
      "n_segments": null,
      "words": null,
      "chars": null,
      "script": null,
      "uwr": null,
      "trailing_3g": null,
      "max_consec_rep": null,
      "classification": "not_run",
      "notes": "Deepgram batch terminated after 3 hung calls; failure mode established on completed clips."
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip06",
      "start_seconds": 992,
      "end_seconds": 1052,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip06.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "elevenlabs",
      "vendor_model": "scribe_v1",
      "language_hint": "som",
      "transcript_present": true,
      "transcript_text": "Kulka muugiyo horaa buka khayr badan yi. Bal mizaanka nabigu sare ah. Wa hal qof. Halqaf iyo qof la xuma qiimeynayo kan horey kulka muugiyiin kaasoo meedana. Midka maaka khayr badan buunabigule alayhi salatu wa salaam. Inkana idinka aan aqteena wax rakaba ahayn ee aleedihin waa fuqradda muslimiintu ka mid yahay. Indallah kaasba kulka muugiiyo kana ka khayr badan buunabigule alayhi salatu wa salaam. Idinku wuxuu hadiisku tilmaamaya mawaaziinta noo hayso naaynaan ay wax sax ahay. Dall'iyarada danbe waxay da'daan u wada iyo xooga iyo sigs baa kaloo muruqyadu dhisayno karaynay wax ku qiimaya. Wani farabuuran. Farabuurandood miyaad tahay fi shari'a. Mizaanka wax lagu qiimayaa ilayh qola munafiqoon tilmaama wuxuu leh. Wayda raa'ytoom tu.",
      "n_segments": 114,
      "words": 114,
      "chars": 741,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.8158,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip06",
      "start_seconds": 992,
      "end_seconds": 1052,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip06.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "whisper",
      "vendor_model": "large-v3 (faster-whisper, local)",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "dhulka muggiyo ki horaa buka xayrbadaan gi bal miizanku nabugu saaray arak wa halqof halqof iyo qof la sumaq qimayna iyo kan horay dhulka muggiyo ka soo midana mitka maa ka xayrbadaan bu nabugu loo ya aleyhi salato wa salam ninkani dhinka aay nakti ina wahbakka mo ahayn ee aalay dihin waa fuqrada muslimintu ka midi ahay inda Allahi ka asbaab dhulka muggiyo ka naka xayrbadaan bu nabugu loo ya aleyhi salato wa salam iyo dhamo ahadid ku tilmaa maya mawaazintanay noo ahaysanay naanay wah saha ahay dhalan iyo rada dhambay wahay dhaa daan uwada iyo xooga iyo wani nsigisbaka loo omur ghiadu li sayyino karabay wahku qimayya wani fara booran fara booran edo miay tahay fi shari'a miizanka wahlogu qimayya ilayhi qola munafiqinu tilmaa mawuhu li",
      "n_segments": 19,
      "words": 127,
      "chars": 743,
      "script": {
        "latin": 1.0,
        "arabic": 0.0,
        "other": 0.0
      },
      "uwr": 0.7323,
      "trailing_3g": 1,
      "max_consec_rep": 2,
      "classification": "surface-usable",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip06",
      "start_seconds": 992,
      "end_seconds": 1052,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip06.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "assemblyai",
      "vendor_model": "universal-2",
      "language_hint": "so",
      "transcript_present": true,
      "transcript_text": "بل ميزان كونه بقو ساره يارك وحلقف حلقف يقفل اسمه قيمين ايوه كان هوره دول كموكيو كاسو ميدانا لماذا لا يوجد خير في النبي عليه الصلاة والسلام؟ لذلك أريد أن أقول لكم ماذا يفعلون فقراء المسلمين؟ عند الله سبحانه وتعالى لا يوجد خير في النبي عليه الصلاة والسلام إذا أردنا أن نتعرف على موازينتنا وحسناينا قوة وحقوقيمية",
      "n_segments": 57,
      "words": 57,
      "chars": 309,
      "script": {
        "latin": 0.0,
        "arabic": 1.0,
        "other": 0.0
      },
      "uwr": 0.8421,
      "trailing_3g": 1,
      "max_consec_rep": 1,
      "classification": "cross-lingual misroute",
      "notes": null
    },
    {
      "video": "salaf_lecture",
      "domain": "religious_lecture",
      "clip_id": "clip06",
      "start_seconds": 992,
      "end_seconds": 1052,
      "duration_seconds": 60,
      "audio_path": "audio/salaf_lecture/clip06.wav",
      "video_url": "https://www.youtube.com/watch?v=qbvgFT5O_d4",
      "vendor": "deepgram",
      "vendor_model": "nova-2",
      "language_hint": "multi (so rejected with 400)",
      "transcript_present": false,
      "transcript_text": null,
      "n_segments": null,
      "words": null,
      "chars": null,
      "script": null,
      "uwr": null,
      "trailing_3g": null,
      "max_consec_rep": null,
      "classification": "not_run",
      "notes": "Deepgram batch terminated after 3 hung calls; failure mode established on completed clips."
    }
  ]
}