{
  "text": "各位小伙伴晚上好。我是初级A。8月15號阿里开源Qwen3.8的27B大語言模型。有點突然但也不意外。而且之前在簽問3.6的時候已經感覺到簽問3.6模型的強大。現在3.8模型也是一個不小的升級。我們簡單來看一下它這裡的介紹。首先簽問3.8模型在代碼。專業工作中都有全面的改進。在推理思維上有高度的可控性。在完成複雜任務更加穩定可靠。\n有更強智能體Agent的執行能力。這次開源的簽問3.8模型是更聰明更強大了。那接下來看一下它的量化模型。簽問的3.8、27B的量化的GGUF模型有許多個版本。大家可以根據自己的顯存大小來選擇模型大小的。比如16G顯存的話,可以選擇Q4、Q3的模型都可以。如果顯存有24G的話,可以選擇Q6、Q5的版本。另外這兩個是它的視覺模型。\n視覺模型等於是它的眼睛可以識別圖像。等於下載一個主模型和一個視覺模型。兩個都是一個視覺模型。選擇其中一個即可。另外簽問3.8、27B模型也有無限制版的。無限制版也是破限和越獄版。它能回答更多受限的問題。大家都懂的。所以無限制版更加符合日常需求。同樣這個版本也有不同大小的。\n我們可以選擇Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、\nQ4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4\n把文件夹名改成Models\n把Qwen3.6的主模型还有视觉模型都放到Models文件中\n然后把下载好的主程序解压好\n解压后的文件重新改名\n文件夹最好不要有中文出现\n我这边改成Llama\n然后再把这个驱动文件再解压出来\n解压后把驱动的这三个文件剪切出来\n再复制到刚刚解压的主程序中\n最后再把Modos文件放到主程序中\n把这两个压缩包删除掉\n接下来我们用ComfyUI做一个启动文件\n首先选择刚才的Lamma文件夹\n直接写\n做一个Lamma快速启动器\n这里的模型\n我们选择是Qwen3或者是DeepSeek\n我选择是Qwen3\n然后直接点发送就行了\n启动器这边已经做好了\n不想自己做的话\n启动器我会放网盘\n大家可以下载\n右键打开启动器的所在的文件夹\n双击启动\n选择主模型\n然后再选择视觉模型\n点启动\n再点打开WebUI\n启动要等一会儿\n因为要加载模型\n然后测试下编程能力\n做一个小游戏\n等了好一会终于搞好\n看来16G的显存\n用16G的模型还是有点吃力的\n要更快的速度的话\n只能用Q3的模型了\n接下来看一下游戏能不能玩\n直接打开游戏\n总体来说\n游戏是没有问题的\n是可以玩的\n游戏的功能\n现在是简单的初级阶段\n可以增加游戏的其他功能或者优化\n比如增加暂停\n优化动态效果等\n由于时间原因\n这里就不继续测试了\n感兴趣的小伙伴\n可以自行部署本地来玩下\n模型程序包和启动器都放网盘了\n有需要的小伙伴可以自行下载\n视频对大家有帮助的话\n来个三连\n谢谢了\n再见了",
  "segments": [
    {
      "id": 0,
      "seek": 0,
      "start": 0.0,
      "end": 29.34,
      "text": "各位小伙伴晚上好。我是初级A。8月15號阿里开源Qwen3.8的27B大語言模型。有點突然但也不意外。而且之前在簽問3.6的時候已經感覺到簽問3.6模型的強大。現在3.8模型也是一個不小的升級。我們簡單來看一下它這裡的介紹。首先簽問3.8模型在代碼。專業工作中都有全面的改進。在推理思維上有高度的可控性。在完成複雜任務更加穩定可靠。",
      "tokens": [
        50365,
        22073,
        7322,
        7384,
        247,
        7384,
        112,
        50157,
        2131,
        1543,
        15914,
        28727,
        16853,
        100,
        32,
        1543,
        23,
        6939,
        5211,
        18616,
        13594,
        15759,
        18937,
        47402,
        18,
        13,
        23,
        1546,
        10076,
        33,
        3582,
        31348,
        12009,
        41908,
        39823,
        1543,
        42440,
        40859,
        5823,
        8395,
        6404,
        1960,
        9042,
        12022,
        1543,
        22942,
        32442,
        3581,
        24579,
        121,
        11361,
        18,
        13,
        21,
        20643,
        25338,
        31742,
        4511,
        24579,
        121,
        11361,
        18,
        13,
        21,
        41908,
        39823,
        1546,
        22752,
        3582,
        1543,
        12648,
        18,
        13,
        23,
        41908,
        39823,
        22021,
        8990,
        1960,
        15686,
        1546,
        41670,
        35315,
        1543,
        5884,
        31995,
        3763,
        28324,
        11284,
        18771,
        1546,
        30312,
        38193,
        1543,
        36490,
        24579,
        121,
        11361,
        18,
        13,
        23,
        41908,
        39823,
        3581,
        19105,
        16337,
        120,
        1543,
        46072,
        27119,
        41315,
        5975,
        48121,
        11319,
        8833,
        1546,
        34490,
        18214,
        1543,
        3581,
        33597,
        13876,
        8870,
        9261,
        255,
        5708,
        2412,
        12979,
        13127,
        1546,
        4429,
        48707,
        21686,
        1543,
        3581,
        41509,
        164,
        97,
        229,
        6306,
        250,
        26443,
        39517,
        19002,
        9990,
        13740,
        102,
        12088,
        4429,
        5363,
        254,
        1543,
        51832
      ],
      "temperature": 0,
      "avg_logprob": -0.10965475266979587,
      "compression_ratio": 1.3154761904761905,
      "no_speech_prob": 1.0479675287278223e-11,
      "text_original": "各位小伙伴晚上好。我是初级A。8月15號阿里开源3.8的27B大語言模型。有點突然但也不意外。而且之前在簽問3.6的時候已經感覺到簽問3.6模型的強大。現在3.8模型也是一個不少的升級。我們簡單來看一下它這裡的介紹。首先簽問3.8模型在代碼。專業工作中都有全面的改進。在推理思維上有高度的可控性。在完成複雜任務更加穩定可靠。",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 1,
      "seek": 2934,
      "start": 29.34,
      "end": 59.18,
      "text": "有更強智能體Agent的執行能力。這次開源的簽問3.8模型是更聰明更強大了。那接下來看一下它的量化模型。簽問的3.8、27B的量化的GGUF模型有許多個版本。大家可以根據自己的顯存大小來選擇模型大小的。比如16G顯存的話,可以選擇Q4、Q3的模型都可以。如果顯存有24G的話,可以選擇Q6、Q5的版本。另外這兩個是它的視覺模型。",
      "tokens": [
        50365,
        2412,
        19002,
        22752,
        5094,
        118,
        8225,
        23987,
        32,
        18594,
        229,
        1546,
        16262,
        115,
        8082,
        8225,
        13486,
        1543,
        2664,
        9487,
        8949,
        47402,
        1546,
        24579,
        121,
        11361,
        18,
        13,
        23,
        41908,
        39823,
        1541,
        19002,
        8171,
        108,
        11100,
        19002,
        22752,
        3582,
        2289,
        1543,
        4184,
        40012,
        28324,
        45224,
        26748,
        23756,
        41908,
        39823,
        1543,
        24579,
        121,
        11361,
        1546,
        18,
        13,
        23,
        1231,
        10076,
        33,
        1546,
        26748,
        23756,
        1546,
        27561,
        52,
        37,
        41908,
        39823,
        2412,
        40914,
        6392,
        3338,
        42096,
        8802,
        1543,
        6868,
        6723,
        31337,
        36841,
        17645,
        1546,
        46605,
        39781,
        3582,
        7322,
        3763,
        21625,
        13167,
        229,
        41908,
        39823,
        3582,
        7322,
        1546,
        1543,
        36757,
        6866,
        38,
        46605,
        39781,
        21358,
        11,
        6723,
        21625,
        13167,
        229,
        48,
        19,
        1231,
        48,
        18,
        1546,
        41908,
        39823,
        7182,
        6723,
        1543,
        13119,
        46605,
        39781,
        2412,
        7911,
        38,
        21358,
        11,
        6723,
        21625,
        13167,
        229,
        48,
        21,
        1231,
        48,
        20,
        1546,
        42096,
        8802,
        1543,
        26202,
        2664,
        34623,
        1541,
        45224,
        27333,
        9463,
        41908,
        39823,
        1543,
        51857
      ],
      "temperature": 0,
      "avg_logprob": -0.05979124915520877,
      "compression_ratio": 1.4436860068259385,
      "no_speech_prob": 1.2418017135418236e-11,
      "text_original": "有更強智能體A震的執行能力。這次開源的簽問3.8模型是更聰明更強大了。那接下來看一下它的量化模型。簽問的3.8、27B的量化的GGUF模型有許多個版本。大家可以根據自己的顯存大小來選擇模型大小的。比如16G顯存的話,可以選擇Q4、Q3的模型都可以。如果顯存有24G的話,可以選擇Q6、Q5的版本。另外這兩個是它的視覺模型。",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 2,
      "seek": 5934,
      "start": 59.34,
      "end": 89.34,
      "text": "視覺模型等於是它的眼睛可以識別圖像。等於下載一個主模型和一個視覺模型。兩個都是一個視覺模型。選擇其中一個即可。另外簽問3.8、27B模型也有無限制版的。無限制版也是破限和越獄版。它能回答更多受限的問題。大家都懂的。所以無限制版更加符合日常需求。同樣這個版本也有不同大小的。",
      "tokens": [
        50365,
        27333,
        9463,
        41908,
        39823,
        10187,
        19488,
        1541,
        45224,
        25281,
        14221,
        249,
        6723,
        43143,
        16158,
        2523,
        244,
        12760,
        1543,
        10187,
        19488,
        4438,
        12290,
        231,
        8990,
        13557,
        41908,
        39823,
        12565,
        8990,
        27333,
        9463,
        41908,
        39823,
        1543,
        34623,
        22796,
        8990,
        27333,
        9463,
        41908,
        39823,
        1543,
        21625,
        13167,
        229,
        9572,
        5975,
        8990,
        39127,
        4429,
        1543,
        26202,
        24579,
        121,
        11361,
        18,
        13,
        23,
        1231,
        10076,
        33,
        41908,
        39823,
        6404,
        2412,
        16976,
        43446,
        25491,
        42096,
        1546,
        1543,
        16976,
        43446,
        25491,
        42096,
        22021,
        39560,
        31699,
        12565,
        25761,
        16262,
        253,
        42096,
        1543,
        11284,
        8225,
        8350,
        28223,
        19002,
        6392,
        23151,
        43446,
        1546,
        17197,
        1543,
        46525,
        29624,
        1546,
        1543,
        7239,
        16976,
        43446,
        25491,
        42096,
        19002,
        9990,
        5437,
        99,
        14245,
        6890,
        11279,
        32535,
        32718,
        1543,
        13089,
        6180,
        6287,
        42096,
        8802,
        6404,
        2412,
        47123,
        3582,
        7322,
        1546,
        1543
      ],
      "temperature": 0,
      "avg_logprob": -0.06514321267604828,
      "compression_ratio": 1.4612546125461254,
      "no_speech_prob": 1.728538666367907e-11,
      "text_original": "視覺模型等於是它的眼睛可以識別圖像。等於下載一個主模型和一個視覺模型。兩個都是一個視覺模型。選擇其中一個即可。另外簽問3.8、27B模型也有無限制版的。無限制版也是破線和越域版。它能回答更多受限的問題。大家都懂的。所以無限制版更加符合日常需求。同樣這個版本也有不同大小的。",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 3,
      "seek": 8934,
      "start": 89.34,
      "end": 119.34,
      "text": "我們可以選擇Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、",
      "tokens": [
        50365,
        5884,
        6723,
        21625,
        13167,
        229,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231
      ],
      "temperature": 0,
      "avg_logprob": -0.20068048040963074,
      "compression_ratio": 10.5,
      "no_speech_prob": 1.841952111669709e-11
    },
    {
      "id": 4,
      "seek": 11934,
      "start": 119.34,
      "end": 149.34,
      "text": "Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4、Q4",
      "tokens": [
        50365,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19,
        1231,
        48,
        19
      ],
      "temperature": 0,
      "avg_logprob": -0.09432554886480084,
      "compression_ratio": 20.38888888888889,
      "no_speech_prob": 7.239950999826661e-11
    },
    {
      "id": 5,
      "seek": 14934,
      "start": 149.34,
      "end": 151.3,
      "text": "把文件夹名改成Models",
      "tokens": [
        50365,
        16075,
        17174,
        20485,
        1787,
        117,
        15940,
        34490,
        11336,
        44,
        378,
        329,
        50463
      ],
      "temperature": 0,
      "avg_logprob": -0.13075345876265546,
      "compression_ratio": 1.4262295081967213,
      "no_speech_prob": 1.0566061914296654e-10,
      "text_original": "把文件夹名改成Modos",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 6,
      "seek": 14934,
      "start": 151.3,
      "end": 156.38,
      "text": "把Qwen3.6的主模型还有视觉模型都放到Models文件中",
      "tokens": [
        50463,
        16075,
        20787,
        23570,
        18,
        13,
        21,
        1546,
        13557,
        41908,
        39823,
        35091,
        40656,
        24447,
        41908,
        39823,
        7182,
        12744,
        4511,
        44,
        378,
        329,
        17174,
        20485,
        5975,
        50717
      ],
      "temperature": 0,
      "avg_logprob": -0.13075345876265546,
      "compression_ratio": 1.4262295081967213,
      "no_speech_prob": 1.0566061914296654e-10,
      "text_original": "把千万3.6的主模型还有视觉模型都放到Modos文件中",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 7,
      "seek": 14934,
      "start": 156.38,
      "end": 160.34,
      "text": "然后把下载好的主程序解压好",
      "tokens": [
        50717,
        26636,
        16075,
        4438,
        17819,
        121,
        20715,
        13557,
        29649,
        6346,
        237,
        17278,
        5014,
        233,
        2131,
        50915
      ],
      "temperature": 0,
      "avg_logprob": -0.13075345876265546,
      "compression_ratio": 1.4262295081967213,
      "no_speech_prob": 1.0566061914296654e-10
    },
    {
      "id": 8,
      "seek": 14934,
      "start": 160.34,
      "end": 164.02,
      "text": "解压后的文件重新改名",
      "tokens": [
        50915,
        17278,
        5014,
        233,
        13547,
        1546,
        17174,
        20485,
        12624,
        12560,
        34490,
        15940,
        51099
      ],
      "temperature": 0,
      "avg_logprob": -0.13075345876265546,
      "compression_ratio": 1.4262295081967213,
      "no_speech_prob": 1.0566061914296654e-10
    },
    {
      "id": 9,
      "seek": 14934,
      "start": 164.02,
      "end": 166.18,
      "text": "文件夹最好不要有中文出现",
      "tokens": [
        51099,
        17174,
        20485,
        1787,
        117,
        8661,
        2131,
        11962,
        2412,
        5975,
        17174,
        7781,
        20204,
        51207
      ],
      "temperature": 0,
      "avg_logprob": -0.13075345876265546,
      "compression_ratio": 1.4262295081967213,
      "no_speech_prob": 1.0566061914296654e-10
    },
    {
      "id": 10,
      "seek": 14934,
      "start": 166.18,
      "end": 167.92000000000002,
      "text": "我这边改成Llama",
      "tokens": [
        51207,
        1654,
        5562,
        40503,
        34490,
        11336,
        43,
        335,
        1696,
        51294
      ],
      "temperature": 0,
      "avg_logprob": -0.13075345876265546,
      "compression_ratio": 1.4262295081967213,
      "no_speech_prob": 1.0566061914296654e-10,
      "text_original": "我这边改成Lamma",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 11,
      "seek": 14934,
      "start": 167.92000000000002,
      "end": 171.22,
      "text": "然后再把这个驱动文件再解压出来",
      "tokens": [
        51294,
        26636,
        8623,
        16075,
        15368,
        24023,
        109,
        34961,
        17174,
        20485,
        8623,
        17278,
        5014,
        233,
        44561,
        51459
      ],
      "temperature": 0,
      "avg_logprob": -0.13075345876265546,
      "compression_ratio": 1.4262295081967213,
      "no_speech_prob": 1.0566061914296654e-10
    },
    {
      "id": 12,
      "seek": 14934,
      "start": 171.22,
      "end": 175.2,
      "text": "解压后把驱动的这三个文件剪切出来",
      "tokens": [
        51459,
        17278,
        5014,
        233,
        13547,
        16075,
        24023,
        109,
        34961,
        1546,
        5562,
        10960,
        7549,
        17174,
        20485,
        5935,
        103,
        23632,
        44561,
        51658
      ],
      "temperature": 0,
      "avg_logprob": -0.13075345876265546,
      "compression_ratio": 1.4262295081967213,
      "no_speech_prob": 1.0566061914296654e-10
    },
    {
      "id": 13,
      "seek": 14934,
      "start": 175.2,
      "end": 178.62,
      "text": "再复制到刚刚解压的主程序中",
      "tokens": [
        51658,
        8623,
        1787,
        235,
        25491,
        4511,
        49160,
        49160,
        17278,
        5014,
        233,
        1546,
        13557,
        29649,
        6346,
        237,
        5975,
        51829
      ],
      "temperature": 0,
      "avg_logprob": -0.13075345876265546,
      "compression_ratio": 1.4262295081967213,
      "no_speech_prob": 1.0566061914296654e-10
    },
    {
      "id": 14,
      "seek": 17934,
      "start": 179.34,
      "end": 183.18,
      "text": "最后再把Modos文件放到主程序中",
      "tokens": [
        50365,
        8661,
        13547,
        8623,
        16075,
        44,
        378,
        329,
        17174,
        20485,
        12744,
        4511,
        13557,
        29649,
        6346,
        237,
        5975,
        50557
      ],
      "temperature": 0,
      "avg_logprob": -0.16395649341268276,
      "compression_ratio": 1.1203703703703705,
      "no_speech_prob": 2.6278023160242192e-11
    },
    {
      "id": 15,
      "seek": 17934,
      "start": 183.18,
      "end": 189.48000000000002,
      "text": "把这两个压缩包删除掉",
      "tokens": [
        50557,
        16075,
        5562,
        36257,
        7549,
        5014,
        233,
        38109,
        102,
        23305,
        2437,
        254,
        32999,
        29327,
        50872
      ],
      "temperature": 0,
      "avg_logprob": -0.16395649341268276,
      "compression_ratio": 1.1203703703703705,
      "no_speech_prob": 2.6278023160242192e-11
    },
    {
      "id": 16,
      "seek": 17934,
      "start": 189.48000000000002,
      "end": 192.9,
      "text": "接下来我们用ComfyUI做一个启动文件",
      "tokens": [
        50872,
        14468,
        4438,
        6912,
        15003,
        9254,
        28846,
        1067,
        10907,
        20182,
        2392,
        107,
        34961,
        17174,
        20485,
        51043
      ],
      "temperature": 0,
      "avg_logprob": -0.16395649341268276,
      "compression_ratio": 1.1203703703703705,
      "no_speech_prob": 2.6278023160242192e-11,
      "text_original": "接下来我们用Workbody做一个启动文件",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 17,
      "seek": 17934,
      "start": 192.9,
      "end": 196.72,
      "text": "首先选择刚才的Lamma文件夹",
      "tokens": [
        51043,
        36490,
        2215,
        231,
        6852,
        102,
        49160,
        18888,
        1546,
        43,
        335,
        1696,
        17174,
        20485,
        1787,
        117,
        51234
      ],
      "temperature": 0,
      "avg_logprob": -0.16395649341268276,
      "compression_ratio": 1.1203703703703705,
      "no_speech_prob": 2.6278023160242192e-11
    },
    {
      "id": 18,
      "seek": 17934,
      "start": 199.72,
      "end": 200.4,
      "text": "直接写",
      "tokens": [
        51384,
        43297,
        5676,
        247,
        51418
      ],
      "temperature": 0,
      "avg_logprob": -0.16395649341268276,
      "compression_ratio": 1.1203703703703705,
      "no_speech_prob": 2.6278023160242192e-11
    },
    {
      "id": 19,
      "seek": 17934,
      "start": 200.4,
      "end": 202.3,
      "text": "做一个Lamma快速启动器",
      "tokens": [
        51418,
        10907,
        20182,
        43,
        335,
        1696,
        10251,
        31217,
        2392,
        107,
        34961,
        34386,
        51513
      ],
      "temperature": 0,
      "avg_logprob": -0.16395649341268276,
      "compression_ratio": 1.1203703703703705,
      "no_speech_prob": 2.6278023160242192e-11
    },
    {
      "id": 20,
      "seek": 17934,
      "start": 202.3,
      "end": 206.12,
      "text": "这里的模型",
      "tokens": [
        51513,
        35102,
        1546,
        41908,
        39823,
        51704
      ],
      "temperature": 0,
      "avg_logprob": -0.16395649341268276,
      "compression_ratio": 1.1203703703703705,
      "no_speech_prob": 2.6278023160242192e-11
    },
    {
      "id": 21,
      "seek": 17934,
      "start": 206.12,
      "end": 208.62,
      "text": "我们选择是Qwen3或者是DeepSeek",
      "tokens": [
        51704,
        15003,
        2215,
        231,
        6852,
        102,
        1541,
        42,
        10121,
        18,
        31148,
        1541,
        11089,
        595,
        50,
        647,
        51829
      ],
      "temperature": 0,
      "avg_logprob": -0.16395649341268276,
      "compression_ratio": 1.1203703703703705,
      "no_speech_prob": 2.6278023160242192e-11,
      "text_original": "我们选择是Kimi3或者是DeepSip",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 22,
      "seek": 20862,
      "start": 208.62,
      "end": 210.96,
      "text": "我选择是Qwen3",
      "tokens": [
        50365,
        1654,
        2215,
        231,
        6852,
        102,
        1541,
        42,
        10121,
        18,
        50482
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11,
      "text_original": "我选择是Kimi3",
      "correction_applied": true,
      "correction_source": "llm_correct_a_presegmentation"
    },
    {
      "id": 23,
      "seek": 20862,
      "start": 210.96,
      "end": 212.86,
      "text": "然后直接点发送就行了",
      "tokens": [
        50482,
        26636,
        43297,
        12579,
        28926,
        29309,
        3111,
        46606,
        50577
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 24,
      "seek": 20862,
      "start": 212.86,
      "end": 214.66,
      "text": "启动器这边已经做好了",
      "tokens": [
        50577,
        2392,
        107,
        34961,
        34386,
        5562,
        40503,
        49161,
        10907,
        12621,
        50667
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 25,
      "seek": 20862,
      "start": 214.66,
      "end": 215.8,
      "text": "不想自己做的话",
      "tokens": [
        50667,
        1960,
        7093,
        17645,
        10907,
        44575,
        50724
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 26,
      "seek": 20862,
      "start": 215.8,
      "end": 217.4,
      "text": "启动器我会放网盘",
      "tokens": [
        50724,
        2392,
        107,
        34961,
        34386,
        1654,
        12949,
        12744,
        16469,
        239,
        5419,
        246,
        50804
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 27,
      "seek": 20862,
      "start": 217.4,
      "end": 218.4,
      "text": "大家可以下载",
      "tokens": [
        50804,
        6868,
        6723,
        4438,
        17819,
        121,
        50854
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 28,
      "seek": 20862,
      "start": 218.4,
      "end": 220.96,
      "text": "右键打开启动器的所在的文件夹",
      "tokens": [
        50854,
        22060,
        23049,
        106,
        12467,
        18937,
        2392,
        107,
        34961,
        34386,
        1546,
        5966,
        3581,
        1546,
        17174,
        20485,
        1787,
        117,
        50982
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 29,
      "seek": 20862,
      "start": 220.96,
      "end": 222.22,
      "text": "双击启动",
      "tokens": [
        50982,
        2129,
        234,
        6336,
        119,
        2392,
        107,
        34961,
        51045
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 30,
      "seek": 20862,
      "start": 222.22,
      "end": 223.92000000000002,
      "text": "选择主模型",
      "tokens": [
        51045,
        2215,
        231,
        6852,
        102,
        13557,
        41908,
        39823,
        51130
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 31,
      "seek": 20862,
      "start": 223.92000000000002,
      "end": 225.78,
      "text": "然后再选择视觉模型",
      "tokens": [
        51130,
        26636,
        8623,
        2215,
        231,
        6852,
        102,
        40656,
        24447,
        41908,
        39823,
        51223
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 32,
      "seek": 20862,
      "start": 225.78,
      "end": 227.94,
      "text": "点启动",
      "tokens": [
        51223,
        12579,
        2392,
        107,
        34961,
        51331
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 33,
      "seek": 20862,
      "start": 227.94,
      "end": 229.46,
      "text": "再点打开WebUI",
      "tokens": [
        51331,
        8623,
        12579,
        12467,
        18937,
        4360,
        65,
        46324,
        51407
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 34,
      "seek": 20862,
      "start": 229.46,
      "end": 230.88,
      "text": "启动要等一会儿",
      "tokens": [
        51407,
        2392,
        107,
        34961,
        4275,
        10187,
        2257,
        12949,
        22933,
        51478
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 35,
      "seek": 20862,
      "start": 230.88,
      "end": 231.92000000000002,
      "text": "因为要加载模型",
      "tokens": [
        51478,
        34627,
        4275,
        9990,
        17819,
        121,
        41908,
        39823,
        51530
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 36,
      "seek": 20862,
      "start": 231.92000000000002,
      "end": 233.76,
      "text": "然后测试下编程能力",
      "tokens": [
        51530,
        26636,
        11038,
        233,
        5233,
        243,
        4438,
        38109,
        244,
        29649,
        8225,
        13486,
        51622
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 37,
      "seek": 20862,
      "start": 233.76,
      "end": 234.68,
      "text": "做一个小游戏",
      "tokens": [
        51622,
        10907,
        20182,
        7322,
        9592,
        116,
        1486,
        237,
        51668
      ],
      "temperature": 0,
      "avg_logprob": -0.11725918839617473,
      "compression_ratio": 1.3120300751879699,
      "no_speech_prob": 1.7243656155740972e-11
    },
    {
      "id": 38,
      "seek": 23862,
      "start": 238.62,
      "end": 242.38,
      "text": "等了好一会终于搞好",
      "tokens": [
        50365,
        10187,
        2289,
        2131,
        2257,
        12949,
        10115,
        230,
        37732,
        41816,
        2131,
        50553
      ],
      "temperature": 0,
      "avg_logprob": -0.1935293197631836,
      "compression_ratio": 1.1518987341772151,
      "no_speech_prob": 2.1282763745800182e-11
    },
    {
      "id": 39,
      "seek": 23862,
      "start": 242.38,
      "end": 243.78,
      "text": "看来16G的显存",
      "tokens": [
        50553,
        4200,
        6912,
        6866,
        38,
        1546,
        1431,
        122,
        39781,
        50623
      ],
      "temperature": 0,
      "avg_logprob": -0.1935293197631836,
      "compression_ratio": 1.1518987341772151,
      "no_speech_prob": 2.1282763745800182e-11
    },
    {
      "id": 40,
      "seek": 23862,
      "start": 243.78,
      "end": 245.9,
      "text": "用16G的模型还是有点吃力的",
      "tokens": [
        50623,
        9254,
        6866,
        38,
        1546,
        41908,
        39823,
        45726,
        2412,
        12579,
        10123,
        13486,
        1546,
        50729
      ],
      "temperature": 0,
      "avg_logprob": -0.1935293197631836,
      "compression_ratio": 1.1518987341772151,
      "no_speech_prob": 2.1282763745800182e-11
    },
    {
      "id": 41,
      "seek": 23862,
      "start": 245.9,
      "end": 247.34,
      "text": "要更快的速度的话",
      "tokens": [
        50729,
        4275,
        19002,
        10251,
        1546,
        31217,
        13127,
        44575,
        50801
      ],
      "temperature": 0,
      "avg_logprob": -0.1935293197631836,
      "compression_ratio": 1.1518987341772151,
      "no_speech_prob": 2.1282763745800182e-11
    },
    {
      "id": 42,
      "seek": 23862,
      "start": 247.34,
      "end": 248.8,
      "text": "只能用Q3的模型了",
      "tokens": [
        50801,
        14003,
        8225,
        9254,
        48,
        18,
        1546,
        41908,
        39823,
        2289,
        50874
      ],
      "temperature": 0,
      "avg_logprob": -0.1935293197631836,
      "compression_ratio": 1.1518987341772151,
      "no_speech_prob": 2.1282763745800182e-11
    },
    {
      "id": 43,
      "seek": 23862,
      "start": 248.8,
      "end": 251.62,
      "text": "接下来看一下游戏能不能玩",
      "tokens": [
        50874,
        14468,
        4438,
        6912,
        28324,
        9592,
        116,
        1486,
        237,
        8225,
        28590,
        19912,
        51015
      ],
      "temperature": 0,
      "avg_logprob": -0.1935293197631836,
      "compression_ratio": 1.1518987341772151,
      "no_speech_prob": 2.1282763745800182e-11
    },
    {
      "id": 44,
      "seek": 23862,
      "start": 251.62,
      "end": 252.62,
      "text": "直接打开游戏",
      "tokens": [
        51015,
        43297,
        12467,
        18937,
        9592,
        116,
        1486,
        237,
        51065
      ],
      "temperature": 0,
      "avg_logprob": -0.1935293197631836,
      "compression_ratio": 1.1518987341772151,
      "no_speech_prob": 2.1282763745800182e-11
    },
    {
      "id": 45,
      "seek": 25262,
      "start": 252.62,
      "end": 263.98,
      "text": "总体来说",
      "tokens": [
        50365,
        33440,
        29485,
        6912,
        8090,
        50933
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 46,
      "seek": 25262,
      "start": 263.98,
      "end": 265.32,
      "text": "游戏是没有问题的",
      "tokens": [
        50933,
        9592,
        116,
        1486,
        237,
        1541,
        17944,
        34069,
        1546,
        51000
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 47,
      "seek": 25262,
      "start": 265.32,
      "end": 266.16,
      "text": "是可以玩的",
      "tokens": [
        51000,
        1541,
        6723,
        19912,
        1546,
        51042
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 48,
      "seek": 25262,
      "start": 266.16,
      "end": 267.42,
      "text": "游戏的功能",
      "tokens": [
        51042,
        9592,
        116,
        1486,
        237,
        1546,
        32311,
        8225,
        51105
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 49,
      "seek": 25262,
      "start": 267.42,
      "end": 269.14,
      "text": "现在是简单的初级阶段",
      "tokens": [
        51105,
        25040,
        1541,
        11249,
        222,
        47446,
        1546,
        28727,
        16853,
        100,
        10034,
        114,
        28427,
        51191
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 50,
      "seek": 25262,
      "start": 269.14,
      "end": 271.84000000000003,
      "text": "可以增加游戏的其他功能或者优化",
      "tokens": [
        51191,
        6723,
        24228,
        252,
        9990,
        9592,
        116,
        1486,
        237,
        1546,
        9572,
        5000,
        32311,
        8225,
        31148,
        7384,
        246,
        23756,
        51326
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 51,
      "seek": 25262,
      "start": 271.84000000000003,
      "end": 273.28000000000003,
      "text": "比如增加暂停",
      "tokens": [
        51326,
        36757,
        24228,
        252,
        9990,
        23826,
        224,
        36135,
        51398
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 52,
      "seek": 25262,
      "start": 273.28000000000003,
      "end": 274.82,
      "text": "优化动态效果等",
      "tokens": [
        51398,
        7384,
        246,
        23756,
        34961,
        3757,
        223,
        43076,
        9319,
        10187,
        51475
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 53,
      "seek": 25262,
      "start": 274.82,
      "end": 276.08,
      "text": "由于时间原因",
      "tokens": [
        51475,
        23786,
        37732,
        44848,
        42577,
        51538
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 54,
      "seek": 25262,
      "start": 276.08,
      "end": 277.52,
      "text": "这里就不继续测试了",
      "tokens": [
        51538,
        35102,
        3111,
        1960,
        10115,
        100,
        10115,
        255,
        11038,
        233,
        5233,
        243,
        2289,
        51610
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 55,
      "seek": 25262,
      "start": 277.52,
      "end": 279.02,
      "text": "感兴趣的小伙伴",
      "tokens": [
        51610,
        9709,
        2347,
        112,
        39835,
        1546,
        7322,
        7384,
        247,
        7384,
        112,
        51685
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 56,
      "seek": 25262,
      "start": 279.02,
      "end": 280.8,
      "text": "可以自行部署本地来玩下",
      "tokens": [
        51685,
        6723,
        9722,
        8082,
        13470,
        16469,
        110,
        8802,
        10928,
        6912,
        19912,
        4438,
        51774
      ],
      "temperature": 0,
      "avg_logprob": -0.08821091397118022,
      "compression_ratio": 1.2567567567567568,
      "no_speech_prob": 1.6327878285538056e-11
    },
    {
      "id": 57,
      "seek": 28080,
      "start": 280.8,
      "end": 283.82,
      "text": "模型程序包和启动器都放网盘了",
      "tokens": [
        50365,
        41908,
        39823,
        29649,
        6346,
        237,
        23305,
        12565,
        2392,
        107,
        34961,
        34386,
        7182,
        12744,
        16469,
        239,
        5419,
        246,
        2289,
        50516
      ],
      "temperature": 0,
      "avg_logprob": -0.13335320306202722,
      "compression_ratio": 1.036764705882353,
      "no_speech_prob": 1.3734257654773874e-11
    },
    {
      "id": 58,
      "seek": 28080,
      "start": 283.82,
      "end": 286.12,
      "text": "有需要的小伙伴可以自行下载",
      "tokens": [
        50516,
        2412,
        35748,
        1546,
        7322,
        7384,
        247,
        7384,
        112,
        6723,
        9722,
        8082,
        4438,
        17819,
        121,
        50631
      ],
      "temperature": 0,
      "avg_logprob": -0.13335320306202722,
      "compression_ratio": 1.036764705882353,
      "no_speech_prob": 1.3734257654773874e-11
    },
    {
      "id": 59,
      "seek": 28080,
      "start": 286.12,
      "end": 288.22,
      "text": "视频对大家有帮助的话",
      "tokens": [
        50631,
        40656,
        39752,
        8713,
        6868,
        2412,
        4845,
        106,
        37618,
        44575,
        50736
      ],
      "temperature": 0,
      "avg_logprob": -0.13335320306202722,
      "compression_ratio": 1.036764705882353,
      "no_speech_prob": 1.3734257654773874e-11
    },
    {
      "id": 60,
      "seek": 28080,
      "start": 288.22,
      "end": 289.1,
      "text": "来个三连",
      "tokens": [
        50736,
        6912,
        7549,
        10960,
        3316,
        252,
        50780
      ],
      "temperature": 0,
      "avg_logprob": -0.13335320306202722,
      "compression_ratio": 1.036764705882353,
      "no_speech_prob": 1.3734257654773874e-11
    },
    {
      "id": 61,
      "seek": 28080,
      "start": 289.1,
      "end": 290.02000000000004,
      "text": "谢谢了",
      "tokens": [
        50780,
        19879,
        2289,
        50826
      ],
      "temperature": 0,
      "avg_logprob": -0.13335320306202722,
      "compression_ratio": 1.036764705882353,
      "no_speech_prob": 1.3734257654773874e-11
    },
    {
      "id": 62,
      "seek": 28080,
      "start": 290.02000000000004,
      "end": 291.04,
      "text": "再见了",
      "tokens": [
        50826,
        44176,
        2289,
        50877
      ],
      "temperature": 0,
      "avg_logprob": -0.13335320306202722,
      "compression_ratio": 1.036764705882353,
      "no_speech_prob": 1.3734257654773874e-11
    }
  ],
  "language": "zh",
  "presegmentation_correction": {
    "applied": true,
    "correction_count": 9,
    "deterministic_fix_count": 0
  }
}