{
  "text": "要說這兩天最火的開源項目,那必定非它莫屬了。不到一周的時間就已拿下1.3萬個Star。\n它由大名鼎鼎的firecrawl 團隊發布,能將任意格式的文檔轉換成markdown,而且還不需依賴第三方服務。速度也快的驚人,僅需幾毫秒。\n線上測試了一下一份100多頁的PDF文件。無任何剪輯加速的情況下,確實眨眼間就轉成了markdown格式的內容。\n可以本地部署也可以為給大模型添加為Source。不管你是自己用,還是集成到agent中,\n甚至拿去開發個小工具或者小插件之類的,它都非常值得研究。",
  "language": "zh",
  "duration": 33.506375,
  "segments": [
    {
      "id": 0,
      "seek": 0,
      "start": 0.0,
      "end": 6.2,
      "text": "要說這兩天最火的開源項目,那必定非它莫屬了。不到一周的時間就已拿下1.3萬個Star。",
      "tokens": [
        50365,
        4275,
        4622,
        2664,
        16313,
        6135,
        8661,
        26171,
        1546,
        8949,
        47402,
        8313,
        227,
        11386,
        11,
        4184,
        28531,
        12088,
        12107,
        11284,
        31127,
        104,
        9636,
        105,
        2289,
        1543,
        35728,
        2257,
        37390,
        1546,
        20788,
        3111,
        15701,
        24351,
        4438,
        16,
        13,
        18,
        14099,
        5155,
        12560,
        46299,
        1543,
        50675
      ],
      "temperature": 0,
      "avg_logprob": -0.13591866996420088,
      "compression_ratio": 1.2091954022988505,
      "no_speech_prob": 8.439612723953882e-12,
      "source": "whisper",
      "text_original": "要說這兩天最火的開源項目,那必定非它莫屬了。不到一周的時間就已拿下1.3萬家新標。",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 1,
      "seek": 0,
      "start": 6.2,
      "end": 15.6,
      "text": "它由大名鼎鼎的firecrawl 團隊發布,能將任意格式的文檔轉換成markdown,而且還不需依賴第三方服務。速度也快的驚人,僅需幾毫秒。",
      "tokens": [
        50675,
        11284,
        23786,
        3582,
        15940,
        41111,
        236,
        41111,
        236,
        1546,
        12037,
        66,
        5131,
        75,
        220,
        37766,
        33929,
        14637,
        34688,
        11,
        8225,
        34168,
        26443,
        9042,
        30921,
        27584,
        1546,
        17174,
        40741,
        242,
        32739,
        36338,
        11336,
        5638,
        5093,
        11,
        22942,
        7824,
        1960,
        32535,
        3254,
        251,
        164,
        7062,
        35878,
        9249,
        27408,
        39517,
        1543,
        31217,
        13127,
        6404,
        10251,
        1546,
        24023,
        248,
        4035,
        11,
        9502,
        227,
        32535,
        23575,
        7256,
        104,
        20956,
        1543,
        51145
      ],
      "temperature": 0,
      "avg_logprob": -0.13591866996420088,
      "compression_ratio": 1.2091954022988505,
      "no_speech_prob": 8.439612723953882e-12,
      "source": "whisper"
    },
    {
      "id": 2,
      "seek": 0,
      "start": 15.6,
      "end": 23.400000000000002,
      "text": "線上測試了一下一份100多頁的PDF文件。無任何剪輯加速的情況下,確實眨眼間就轉成了markdown格式的內容。",
      "tokens": [
        51145,
        31699,
        5708,
        9592,
        105,
        22099,
        2289,
        8861,
        2257,
        36266,
        6879,
        6392,
        8313,
        223,
        1546,
        17349,
        37,
        17174,
        20485,
        1543,
        16976,
        26443,
        11906,
        5935,
        103,
        12290,
        107,
        9990,
        31217,
        1546,
        39391,
        4438,
        11,
        24293,
        10376,
        2862,
        101,
        25281,
        11016,
        3111,
        32739,
        11336,
        2289,
        5638,
        5093,
        30921,
        27584,
        1546,
        28472,
        25750,
        1543,
        51535
      ],
      "temperature": 0,
      "avg_logprob": -0.13591866996420088,
      "compression_ratio": 1.2091954022988505,
      "no_speech_prob": 8.439612723953882e-12,
      "source": "whisper"
    },
    {
      "id": 3,
      "seek": 0,
      "start": 23.400000000000002,
      "end": 29.04,
      "text": "可以本地部署也可以為給大模型添加為Source。不管你是自己用,還是集成到agent中,",
      "tokens": [
        51535,
        6723,
        8802,
        10928,
        13470,
        16469,
        110,
        47935,
        6344,
        17798,
        3582,
        41908,
        39823,
        14176,
        119,
        9990,
        6344,
        16806,
        1220,
        1543,
        45376,
        32526,
        17645,
        9254,
        11,
        25583,
        26020,
        11336,
        4511,
        559,
        317,
        5975,
        11,
        51817
      ],
      "temperature": 0,
      "avg_logprob": -0.13591866996420088,
      "compression_ratio": 1.2091954022988505,
      "no_speech_prob": 8.439612723953882e-12,
      "source": "whisper",
      "text_original": "可以本地部署也可以為給大模型添加為Scale。不管你是自己用,還是集成到agent中,",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 4,
      "seek": 2904,
      "start": 29.04,
      "end": 33.44,
      "text": "甚至拿去開發個小工具或者小插件之類的,它都非常值得研究。",
      "tokens": [
        50365,
        17556,
        20844,
        24351,
        6734,
        8949,
        14637,
        3338,
        7322,
        23323,
        39806,
        31148,
        7322,
        11673,
        240,
        20485,
        9574,
        40092,
        1546,
        11,
        11284,
        7182,
        14392,
        40242,
        5916,
        23230,
        242,
        44704,
        1543,
        50585
      ],
      "temperature": 0,
      "avg_logprob": -0.07986680922969695,
      "compression_ratio": 0.8817204301075269,
      "no_speech_prob": 2.0872104392055668e-11,
      "source": "whisper"
    }
  ],
  "transcript_selection": {
    "schema": "videonote.transcript_selection.v1",
    "status": "SELECTED",
    "requested_language": "zh",
    "duration": 33.506375,
    "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": 5,
      "multilingual": false,
      "first_start": 0.0,
      "final_end": 33.44,
      "duration_coverage": 0.998019,
      "maximum_gap_seconds": 0.0,
      "inverted_count": 0,
      "language": {
        "requested": "zh",
        "matched": true,
        "latin": 38,
        "han": 177,
        "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": 33.506,
      "rejected_segment_count": 0,
      "rejected_segments": []
    }
  },
  "presegmentation_correction": {
    "applied": true,
    "correction_count": 2,
    "deterministic_fix_count": 0
  }
}