{
  "text": "大家好。\n今日AI分享Audio8 ASR 0.1B，僅0.1B參數。\n超輕量多語言語音識別模型來了。\n準備好項目文件並完成解壓。\n雙擊Run.bat，頁面搜索127.0.0.1:7860。\n上傳音頻文件,修改推理參數,然後點擊開始識別按鈕。\n兩年半的個人練習生蔡徐坤,喜歡唱、跳、rap、籃球、music。\n這裡識別的內容有問題,可以通過熱詞控制。\n把 '唱、跳、rap、籃球、music' 改成 '唱、跳、rap、籃球、音樂'。\n最後再次點擊生成按鈕,就可以開始創作了。\n如果覺得這個項目不錯,記得點贊關注。\n後續繼續分享更多AI開源項目。\n最後再次點擊生成按鈕，就可以開始識別了。\n如果覺得這個項目不錯，記得點贊關注。\n後續繼續分享更多AI開源項目。",
  "language": "zh",
  "duration": 83.150688,
  "segments": [
    {
      "id": 0,
      "seek": 0,
      "start": 0.0,
      "end": 1.0,
      "text": "大家好。",
      "tokens": [
        50365,
        15248,
        1543,
        50415
      ],
      "temperature": 0,
      "avg_logprob": -0.2627332936162534,
      "compression_ratio": 1.0489795918367346,
      "no_speech_prob": 1.1377490963249137e-11,
      "source": "whisper"
    },
    {
      "id": 1,
      "seek": 0,
      "start": 1.0,
      "end": 7.0,
      "text": "今日AI分享Audio8 ASR 0.1B，僅0.1B參數。",
      "tokens": [
        50415,
        26259,
        48698,
        30855,
        15591,
        1004,
        38666,
        7469,
        49,
        15,
        13,
        16,
        33,
        11,
        9502,
        227,
        15,
        13,
        16,
        33,
        47375,
        30622,
        1543,
        50715
      ],
      "temperature": 0,
      "avg_logprob": -0.2627332936162534,
      "compression_ratio": 1.0489795918367346,
      "no_speech_prob": 1.1377490963249137e-11,
      "source": "whisper",
      "text_original": "今日AI分享AudioBar ASR0.1B,僅0.1B參數。",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 2,
      "seek": 0,
      "start": 7.0,
      "end": 10.0,
      "text": "超輕量多語言語音識別模型來了。",
      "tokens": [
        50715,
        19869,
        34801,
        26748,
        6392,
        31348,
        12009,
        31348,
        18034,
        43143,
        16158,
        41908,
        39823,
        25364,
        1543,
        50865
      ],
      "temperature": 0,
      "avg_logprob": -0.2627332936162534,
      "compression_ratio": 1.0489795918367346,
      "no_speech_prob": 1.1377490963249137e-11,
      "source": "whisper"
    },
    {
      "id": 3,
      "seek": 0,
      "start": 10.0,
      "end": 13.0,
      "text": "準備好項目文件並完成解壓。",
      "tokens": [
        50865,
        36241,
        2131,
        8313,
        227,
        11386,
        17174,
        20485,
        35934,
        41509,
        17278,
        45964,
        1543,
        51015
      ],
      "temperature": 0,
      "avg_logprob": -0.2627332936162534,
      "compression_ratio": 1.0489795918367346,
      "no_speech_prob": 1.1377490963249137e-11,
      "source": "whisper"
    },
    {
      "id": 4,
      "seek": 0,
      "start": 13.0,
      "end": 20.0,
      "text": "雙擊Run.bat，頁面搜索127.0.0.1:7860。",
      "tokens": [
        51015,
        6306,
        247,
        37813,
        42639,
        11,
        33,
        302,
        11,
        8313,
        223,
        8833,
        162,
        14987,
        7732,
        95,
        4762,
        22,
        13,
        15,
        13,
        15,
        13,
        16,
        11,
        30693,
        4550,
        1543,
        51365
      ],
      "temperature": 0,
      "avg_logprob": -0.2627332936162534,
      "compression_ratio": 1.0489795918367346,
      "no_speech_prob": 1.1377490963249137e-11,
      "source": "whisper",
      "text_original": "雙擊Run,Bet,頁面搜索127.0.0.1,7860。",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 5,
      "seek": 0,
      "start": 24.0,
      "end": 29.0,
      "text": "上傳音頻文件,修改推理參數,然後點擊開始識別按鈕。",
      "tokens": [
        51565,
        5708,
        40852,
        18034,
        29940,
        17174,
        20485,
        11,
        7792,
        106,
        34490,
        33597,
        13876,
        47375,
        30622,
        11,
        10213,
        8216,
        37813,
        21017,
        43143,
        16158,
        26613,
        41597,
        243,
        1543,
        51815
      ],
      "temperature": 0,
      "avg_logprob": -0.2627332936162534,
      "compression_ratio": 1.0489795918367346,
      "no_speech_prob": 1.1377490963249137e-11,
      "source": "whisper"
    },
    {
      "id": 6,
      "seek": 3000,
      "start": 30.0,
      "end": 35.0,
      "text": "兩年半的個人練習生蔡徐坤,喜歡唱、跳、rap、籃球、music。",
      "tokens": [
        50365,
        16313,
        5157,
        30018,
        1546,
        40104,
        47027,
        34025,
        8244,
        41242,
        94,
        2172,
        238,
        14872,
        97,
        11,
        21385,
        39861,
        1231,
        29366,
        1231,
        4007,
        1231,
        22113,
        225,
        28533,
        1231,
        7572,
        1543,
        50615
      ],
      "temperature": 0,
      "avg_logprob": -0.15663105249404907,
      "compression_ratio": 1.3396226415094339,
      "no_speech_prob": 1.7921168021794998e-11,
      "source": "whisper"
    },
    {
      "id": 7,
      "seek": 3000,
      "start": 41.0,
      "end": 44.0,
      "text": "這裡識別的內容有問題,可以通過熱詞控制。",
      "tokens": [
        50915,
        18771,
        43143,
        16158,
        1546,
        28472,
        25750,
        2412,
        17197,
        11,
        6723,
        19550,
        8816,
        32745,
        6716,
        252,
        48707,
        25491,
        1543,
        51065
      ],
      "temperature": 0,
      "avg_logprob": -0.15663105249404907,
      "compression_ratio": 1.3396226415094339,
      "no_speech_prob": 1.7921168021794998e-11,
      "source": "whisper"
    },
    {
      "id": 8,
      "seek": 3000,
      "start": 50.0,
      "end": 57.0,
      "text": "把 '唱、跳、rap、籃球、music' 改成 '唱、跳、rap、籃球、音樂'。",
      "tokens": [
        51365,
        16075,
        39861,
        1231,
        29366,
        1231,
        4007,
        1231,
        22113,
        225,
        28533,
        1231,
        7572,
        11,
        34490,
        11336,
        39861,
        1231,
        29366,
        1231,
        4007,
        1231,
        22113,
        225,
        28533,
        1231,
        7572,
        1543,
        51715
      ],
      "temperature": 0,
      "avg_logprob": -0.15663105249404907,
      "compression_ratio": 1.3396226415094339,
      "no_speech_prob": 1.7921168021794998e-11,
      "source": "whisper",
      "text_original": "把唱、跳、rap、籃球、music,改成唱、跳、rap、籃球、music。",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 9,
      "seek": 6000,
      "start": 60.0,
      "end": 63.0,
      "text": "最後再次點擊生成按鈕,就可以開始創作了。",
      "tokens": [
        50365,
        20578,
        8623,
        9487,
        8216,
        37813,
        8244,
        11336,
        26613,
        41597,
        243,
        11,
        36781,
        21017,
        5935,
        113,
        11914,
        2289,
        1543,
        50515
      ],
      "temperature": 0,
      "avg_logprob": -0.17646134341204608,
      "compression_ratio": 1.0134228187919463,
      "no_speech_prob": 3.3218178208116456e-11,
      "source": "whisper"
    },
    {
      "id": 10,
      "seek": 6000,
      "start": 63.0,
      "end": 64.0,
      "text": "如果覺得這個項目不錯,記得點贊關注。",
      "tokens": [
        50515,
        13119,
        11274,
        6287,
        8313,
        227,
        11386,
        30967,
        11,
        30872,
        8216,
        18464,
        232,
        14899,
        26432,
        1543,
        50565
      ],
      "temperature": 0,
      "avg_logprob": -0.17646134341204608,
      "compression_ratio": 1.0134228187919463,
      "no_speech_prob": 3.3218178208116456e-11,
      "source": "whisper"
    },
    {
      "id": 11,
      "seek": 6000,
      "start": 64.0,
      "end": 66.0,
      "text": "後續繼續分享更多AI開源項目。",
      "tokens": [
        50565,
        5661,
        29583,
        38459,
        30855,
        19002,
        6392,
        48698,
        8949,
        47402,
        8313,
        227,
        11386,
        1543,
        50665
      ],
      "temperature": 0,
      "avg_logprob": -0.17646134341204608,
      "compression_ratio": 1.0134228187919463,
      "no_speech_prob": 3.3218178208116456e-11,
      "source": "whisper"
    },
    {
      "id": 12,
      "seek": 6600,
      "start": 66.0,
      "end": 71.0,
      "text": "最後再次點擊生成按鈕，就可以開始識別了。",
      "tokens": [
        50365,
        20578,
        8623,
        9487,
        8216,
        37813,
        8244,
        11336,
        26613,
        41597,
        243,
        11,
        36781,
        21017,
        5935,
        113,
        11914,
        2289,
        1543,
        50615
      ],
      "temperature": 0,
      "avg_logprob": -0.1433346977940312,
      "compression_ratio": 1.0134228187919463,
      "no_speech_prob": 3.148125857221906e-11,
      "source": "whisper",
      "text_original": "最後再次點擊生成按鈕,就可以開始創作了。",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 13,
      "seek": 6600,
      "start": 71.0,
      "end": 72.0,
      "text": "如果覺得這個項目不錯，記得點贊關注。",
      "tokens": [
        50615,
        13119,
        11274,
        6287,
        8313,
        227,
        11386,
        30967,
        11,
        30872,
        8216,
        18464,
        232,
        14899,
        26432,
        1543,
        50665
      ],
      "temperature": 0,
      "avg_logprob": -0.1433346977940312,
      "compression_ratio": 1.0134228187919463,
      "no_speech_prob": 3.148125857221906e-11,
      "source": "whisper",
      "text_original": "如果覺得這個項目不錯,記得點贊關注。",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 14,
      "seek": 6600,
      "start": 72.0,
      "end": 76.0,
      "text": "後續繼續分享更多AI開源項目。",
      "tokens": [
        50665,
        5661,
        29583,
        38459,
        30855,
        19002,
        6392,
        48698,
        8949,
        47402,
        8313,
        227,
        11386,
        1543,
        50865
      ],
      "temperature": 0,
      "avg_logprob": -0.1433346977940312,
      "compression_ratio": 1.0134228187919463,
      "no_speech_prob": 3.148125857221906e-11,
      "source": "whisper"
    }
  ],
  "transcript_selection": {
    "schema": "videonote.transcript_selection.v1",
    "status": "SELECTED",
    "requested_language": "zh",
    "duration": 83.150688,
    "reference_path": null,
    "reference_kind": null,
    "reference_language": null,
    "selected_source": "whisper_segmented",
    "decision_reason": "source_unavailable_or_untrusted_whisper_passed",
    "whisper_rejected": false,
    "source_profile": {
      "segment_count": 0,
      "duration_coverage": 0.0,
      "usable_full": false,
      "usable_partial": false,
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "language": {
        "requested": "zh",
        "matched": false
      }
    },
    "whisper_profile": {
      "segment_count": 15,
      "multilingual": false,
      "first_start": 0.0,
      "final_end": 76.0,
      "duration_coverage": 0.914003,
      "maximum_gap_seconds": 6.0,
      "inverted_count": 0,
      "language": {
        "requested": "zh",
        "matched": true,
        "latin": 49,
        "han": 203,
        "kana": 0,
        "hangul": 0
      },
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "usable_full": true,
      "usable_partial": true
    },
    "comparison": {
      "window_seconds": 30.0,
      "compared_window_count": 0,
      "median_score": 0.0,
      "divergent_window_count": 0,
      "divergent_ratio": 0.0,
      "divergent_windows": []
    },
    "timestamp_alignment": {
      "applied": false,
      "estimated_offset_seconds": 0.0,
      "matched_segment_count": 0,
      "match_scores": []
    },
    "caption_timing_repair": null,
    "filled_source_profile": null,
    "whisper_duration_guard": {
      "media_duration": 83.151,
      "rejected_segment_count": 0,
      "rejected_segments": []
    }
  },
  "presegmentation_correction": {
    "applied": true,
    "correction_count": 5,
    "deterministic_fix_count": 0
  }
}