{ "text": "大家好,歡迎回來。今天給大家分享Subflow,是我們公司開發的一款AI自動識別語音。\n然後進行AI的配音,以及視頻字幕的一個翻譯和視頻字幕的自動生成與燒錄的一款工具。\nSubflow目前支持了Windows版本和MacOS的一個版本,大家可以去下載到完整的版本。\n我們把整個項目已經開源了。那我們看一下這個項目。進到這個項目裡面來是這個樣式。\n我們可以在這個地方去上傳本地的視頻。也可以在這個地方下載YouTube或者Bilibili的視頻。\n然後我們可以針對我們的視頻字幕進行處理。如果你的原視頻是沒有任何字幕的,那我們可以用這款工具加上字幕。\n那這裡還有就是如果原視頻它已經有了音頻對吧。\n那有一個音頻的情況下,比如你想把它的語言,比如從中文換成英文對吧。\n那它就會自動調用千問的配音的模型,然後去選擇一種音色。\n然後我們就可以進行一個配音,然後生成中英文的字幕,然後燒入到整個系統裡面來。\n那這裡面首先我們調用了配音或者說音色的這個模型,是謙問的TTS的一個開源模型。\n大家可以在本地的算力上進行運行,也是免費的。\n那除了之外呢也是支持了你自己用自己的原聲進行克隆。\n直接用你自己的聲音然後去發出相關聲音來。\n這裡的標準音色比如你是個中文的人,那你說英文那你的標準音色。\n配下來沒有我們這種標準音色做的那麼流暢吧。\n所以說這裡會有兩個版本,一個是標準音色版本,一個是克隆原聲的一個版本。\n當然還有一個GPT的Soviet,用這個Soviet也可以去,有一些語音也是可以的。\n但是這裡的那個語音與音色裡面就會更少一些。\n那我這裡推薦大家用這個QN3TTS,如果你要用自己的語音,那你就用QN3TTS克隆的原聲。\n好這是第一個關於配音和音色的一個選擇。\n那此外呢我們還可以選擇字幕的一個顏色。比如我這裡選了中音的字幕,或者說音中的字幕。\n中音的字幕就是說中文在上面,英文在下面,那反過來就是英文在上面和中文在下面。\n那也支持其他語種,包括什麼德語、法語、日語、繁體中文都是沒有問題。\n那字幕的顏色大家也可以看到,比如我們這裡有中文和英文字幕的一個顏色,我們可以在這裡進行選擇。\n比如你中文要一個顏色,英文要一個顏色也是可以的。\n好,講到字幕的顏色,除此之外呢還有這個叫電影級的一個潤色。\n可以把你的字幕做一些深度的AI校對,以及做相關的一個潤色處理,會表現更好一些。\n這個也是推薦大家去勾上的。但這個勾上的話,它調用的就是我們API站內的大語言模型。\n我們可以看一下這裡有一個語音識別和翻譯功能。\n那翻譯的話其實是需要用到我們的一個API的。\n那我們也可以直接進到我們的API分發站裡面來,然後選擇自己的一個令牌,點擊控制台。\n然後API密鑰。我們複製自己的這個令牌。\n然後填入到這個框框裡面來。填入之後我們點擊保存令牌。\n然後獲取模型。\n比如我在這裡選一個5.6的terra。\n或者說選一個luna。用來跑翻譯還是很方便的嘛。\n那我就選這個terra了。然後OK,我們現在就已經全部準備好了。我們視頻也已經上傳了。對吧。\n或者說你直接通過鏈接去下載的。也可以。你直接鏈接下載之後,你那個視頻也會自動的加載過來。\n就不需要你二次上傳了。好,那我們處理完成之後,我們只要點一下開始處理這個按鈕就可以。\n我們點擊開始處理。那這裡會自動的進行配音環境的一個校對。如果你缺相關的環境,它也會自動的幫你去裝。\n那這個其實這個後戶端,我已經幫大家裝好了那個配音的環境。然後還有千問3的這個模型也給大家配好了。\n還有這個語音識別的whisper的模型。這裡一點需要大家需要注意。如果你本地有顯卡的話,你可以選這個median。\n如果你本地沒有顯卡的話,你可以選這個small或者說tiny去做語音的識別。\n這個是需要本地算力的,包括上面這個千問3這個TTS本質都是算力的。那我可以給大家看一下。\n像我任務管理器裡面的一些東西。任務管理器點擊性能。然後大家可以看到這裡的英偉達顯卡是會直接去占用的。\n特別是macOS如果你是8GB的內存,你就不要想了。這個你基本跑不了。\n你至少也要有個16G或者說24G去運行,會比較流暢一些。而這個整個客戶端會比較大一點。大家需要注意,它大概是有10G\n一個Gb左右吧。大概這麼大。因為裡面有一些模型。然後還有就是除了說模型,還有就是一些依賴環境。\nPython的一些依賴環境。包括我們的配音模型和語音識別的這兩個模型。本質都在本地跑。你的這個雲端API只是用來跑那個翻譯的。\n我覺得接一個API去跑翻譯,效果會更好一些。等這個流程跑完。我們去看一下它的效果就好。\n好,我們來看一下這個任務快完成了。首先它做的是這個音頻的抽取。抽取完了音頻之後,進行語音的一個識別。\n然後整理了我們的一個字幕。然後再將字幕做了一個翻譯。因為我們的目標語言是英文。\n翻譯成了英語之後,再進行一個TTS的一個配音。配音完成之後,最後就是合成我們的音頻。\n那這次已經進行到了最後這一個片段,就是分成了70個片段嘛。那這裡就是具體的一個工作流程。\n整個操作其實就很簡單。你上傳的視頻,然後選好你的目標語種。對吧。還有一些字幕的顏色。或者說電影集的一些潤色。\n然後配一個那個翻譯的模型。那就在這裡。我這裡選了一個翻譯的模型。最後就是點一下開始處理。\n然後這裡就會進行自動化的處理。現在再進行混合音軌,並且封裝到這個視頻裡面來。\n那我們可以點開去看一下。到時候應該就會在我們這一個Subflow的,\n這個項目裡面。那這裡會生成一個視頻。這裡是剛才的這一個視頻。\n這是原版的視頻,我們可以先打開來看一眼。\n然後通過這個。\n這是原版的這個視頻。\n然後它是完全沒有字幕的。是我之前自己錄的一個視頻。\n對吧。然後完全沒有字幕。那這裡生成的是這個音中版本的。\n那這裡還沒有。這裡已經生成完成了。我們可以看一下。\n打開這個。\n打開這個。\n音中字幕的一個版本。那下面兩個是它的那個字幕。\nSRT的一個字幕。那我們打開這個視頻。這個視頻已經燒錄好了。就是把字幕那些東西都加進去了。\n大家可以看到。這裡不只是說。\n我們把這個中文的視頻換成了一個英文的視頻。\n而且配上了英文加中文的一個字幕。那這裡的音色。我用的是。\n千文的TTS的一個音色。然後選的是第一個男聲的一個聲音。OK。那這是。\n這一個演示。那大家可以看一下。就每一塊。\n大家都配。配上了相關的音中字幕。\n大家一般用的比較多的還是。\n國內小夥伴。一般用的比較多的就是這個中音字幕。\n就是中音雙語字幕。然後目標語種是中文。那他就不需要這個配音。\n和這個選音色。因為就是用你自己的一個聲音。\n那只有你選擇目標語種的時候。他才會出來配音。\nOK。這是一塊。第二塊就是說。\n當大家在做這個subflow。特別是。\n在進行這個視頻下的時候。你的那個視頻的那個cookie可能會到期。\n你需要更換自己的cookie。那怎麼去更換呢?\n是需要在這個cookies的項目文件夾內有兩個文件。\n一個是youtube的cookie。一個是bilibili的cookies。\n那我們怎麼去獲取這個cookie呢?\n首先第一步我們是需要在那個我們的導航頁。然後往下滑。\n往下找。找到有一個叫chrome插件的商店。當然你用的edg。\n你可以用這個edg的插件商店。然後進到這裡面來之後。\n我們搜索getcookies。搜索getcookies。大家可以看到我安裝的是這個版本。\n叫getcookies.textclean。這個東西。那怎麼去用呢?\n比如我這裡是youtube。那我只要在這個位置點一下。\n然後我們就可以copy他的youtube的cookies。\n那我們把這個cookies複製一下。然後填到這一個文檔裡面來。\n然後雙擊打開。那填到這裡面來。把原來這個刪除。\n如果你沒有的話。那你只貼進來。然後保存上。\n這個時候我們的cookies就更換成功了。\nYouTube。Bilibili是一樣的。你只要隨機打開Bilibili的網頁。\n然後用同樣的方法。比如我這裡隨便打開一個Bilibili的網站。\n然後我只要點一下這個按鈕。然後點擊copy。\n就可以把b站的cookies給他copy下來。然後這樣的話就可以再線去把雲端的視頻給他下載下來。\n當然如果你是本地視頻你直接上傳就好了。\n有關於我們這一款叫Subflow語幕的這一款。\n直接提供了我們Windows和MacOS的一個客戶端。大家直接下載客戶端去使用就好了。\n好。大家對這個項目感興趣的話可以幫我們點一點小信心。點一下fork。\n謝謝。這就是本期視頻的全部內容。謝謝大家。", "language": "zh", "duration": 597.472688, "segments": [ { "id": 0, "seek": 0, "start": 0.0, "end": 9.0, "text": "大家好,歡迎回來。今天給大家分享Subflow,是我們公司開發的一款AI自動識別語音。", "tokens": [ 50365, 15248, 11, 21315, 46556, 1543, 12074, 17798, 6868, 30855, 39582, 10565, 11, 1541, 5884, 13545, 32981, 8949, 14637, 1546, 2257, 48798, 48698, 9722, 12572, 43143, 16158, 31348, 18034, 1543, 50815 ], "temperature": 0, "avg_logprob": -0.17567602193580484, "compression_ratio": 1.2196969696969697, "no_speech_prob": 8.232803327956617e-12, "source": "whisper" }, { "id": 1, "seek": 0, "start": 9.0, "end": 16.0, "text": "然後進行AI的配音,以及視頻字幕的一個翻譯和視頻字幕的自動生成與燒錄的一款工具。", "tokens": [ 50815, 10213, 18214, 8082, 48698, 1546, 38846, 18034, 11, 40282, 27333, 29940, 22381, 40211, 1546, 8990, 42716, 11823, 107, 12565, 27333, 29940, 22381, 40211, 1546, 9722, 12572, 8244, 11336, 34127, 38168, 115, 9567, 226, 1546, 2257, 48798, 23323, 39806, 1543, 51165 ], "temperature": 0, "avg_logprob": -0.17567602193580484, "compression_ratio": 1.2196969696969697, "no_speech_prob": 8.232803327956617e-12, "source": "whisper", "text_original": "然後進行AI的配音,以及視頻字幕的一個翻譯和視頻字幕的自動生成與騷錄的一款工具。", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "id": 2, "seek": 0, "start": 16.0, "end": 23.0, "text": "Subflow目前支持了Windows版本和MacOS的一個版本,大家可以去下載到完整的版本。", "tokens": [ 51165, 39582, 10565, 39004, 35488, 2289, 45813, 1509, 42096, 8802, 12565, 34355, 4367, 1546, 8990, 42096, 8802, 11, 6868, 6723, 6734, 4438, 12290, 231, 4511, 14128, 27662, 1546, 42096, 8802, 1543, 51515 ], "temperature": 0, "avg_logprob": -0.17567602193580484, "compression_ratio": 1.2196969696969697, "no_speech_prob": 8.232803327956617e-12, "source": "whisper" }, { "id": 3, "seek": 2300, "start": 23.0, "end": 31.0, "text": "我們把整個項目已經開源了。那我們看一下這個項目。進到這個項目裡面來是這個樣式。", "tokens": [ 50365, 5884, 16075, 27662, 3338, 8313, 227, 11386, 25338, 8949, 2523, 241, 2289, 1543, 46714, 28324, 6287, 8313, 227, 11386, 1543, 18214, 4511, 6287, 8313, 227, 11386, 32399, 3763, 1541, 6287, 6180, 27584, 1543, 50765 ], "temperature": 0, "avg_logprob": -0.19140220234412272, "compression_ratio": 1.508361204013378, "no_speech_prob": 1.1709425096206871e-11, "source": "whisper", "text_original": "我們把整個項目已經開圓了。那我們看一下這個項目。進到這個項目裡面來是這個樣式。", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "id": 4, "seek": 2300, "start": 31.0, "end": 39.0, "text": "我們可以在這個地方去上傳本地的視頻。也可以在這個地方下載YouTube或者Bilibili的視頻。", "tokens": [ 50765, 5884, 6723, 3581, 6287, 30146, 6734, 5708, 40852, 8802, 10928, 1546, 27333, 29940, 1543, 47935, 3581, 6287, 30146, 4438, 12290, 231, 3223, 41173, 31148, 33, 12722, 2312, 1546, 27333, 29940, 1543, 51165 ], "temperature": 0, "avg_logprob": -0.19140220234412272, "compression_ratio": 1.508361204013378, "no_speech_prob": 1.1709425096206871e-11, "source": "whisper" }, { "id": 5, "seek": 2300, "start": 39.0, "end": 47.0, "text": "然後我們可以針對我們的視頻字幕進行處理。如果你的原視頻是沒有任何字幕的,那我們可以用這款工具加上字幕。", "tokens": [ 51165, 10213, 5884, 6723, 5873, 251, 2855, 33641, 27333, 29940, 22381, 40211, 18214, 8082, 39289, 13876, 1543, 13119, 18961, 19683, 27333, 29940, 1541, 6963, 26443, 11906, 22381, 40211, 1546, 11, 46714, 6723, 9254, 2664, 48798, 23323, 39806, 9990, 5708, 22381, 40211, 1543, 51565 ], "temperature": 0, "avg_logprob": -0.19140220234412272, "compression_ratio": 1.508361204013378, "no_speech_prob": 1.1709425096206871e-11, "source": "whisper" }, { "id": 6, "seek": 2300, "start": 47.0, "end": 51.0, "text": "那這裡還有就是如果原視頻它已經有了音頻對吧。", "tokens": [ 51565, 4184, 18771, 15569, 5620, 13119, 19683, 27333, 29940, 11284, 25338, 2412, 2289, 18034, 29940, 47905, 1543, 51765 ], "temperature": 0, "avg_logprob": -0.19140220234412272, "compression_ratio": 1.508361204013378, "no_speech_prob": 1.1709425096206871e-11, "source": "whisper" }, { "id": 7, "seek": 5100, "start": 51.0, "end": 56.0, "text": "那有一個音頻的情況下,比如你想把它的語言,比如從中文換成英文對吧。", "tokens": [ 50365, 4184, 2412, 8990, 18034, 29940, 1546, 39391, 4438, 11, 36757, 38386, 42061, 1546, 31348, 12009, 11, 36757, 21068, 5975, 17174, 36338, 11336, 27869, 17174, 47905, 1543, 50615 ], "temperature": 0, "avg_logprob": -0.11461719306739601, "compression_ratio": 1.4353312302839116, "no_speech_prob": 1.8985370567325965e-11, "source": "whisper" }, { "id": 8, "seek": 5100, "start": 56.0, "end": 62.0, "text": "那它就會自動調用千問的配音的模型,然後去選擇一種音色。", "tokens": [ 50615, 4184, 11284, 47778, 9722, 12572, 32943, 9254, 48, 17174, 1546, 38846, 18034, 1546, 41908, 39823, 11, 10213, 6734, 21625, 13167, 229, 2257, 16516, 18034, 17673, 1543, 50915 ], "temperature": 0, "avg_logprob": -0.11461719306739601, "compression_ratio": 1.4353312302839116, "no_speech_prob": 1.8985370567325965e-11, "source": "whisper", "text_original": "那它就會自動調用Q文的配音的模型,然後去選擇一種音色。", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "id": 9, "seek": 5100, "start": 62.0, "end": 68.0, "text": "然後我們就可以進行一個配音,然後生成中英文的字幕,然後燒入到整個系統裡面來。", "tokens": [ 50915, 10213, 5884, 36781, 18214, 8082, 8990, 38846, 18034, 11, 10213, 8244, 11336, 5975, 27869, 17174, 1546, 22381, 40211, 11, 10213, 24184, 240, 14028, 4511, 27662, 3338, 25368, 33725, 32399, 3763, 1543, 51215 ], "temperature": 0, "avg_logprob": -0.11461719306739601, "compression_ratio": 1.4353312302839116, "no_speech_prob": 1.8985370567325965e-11, "source": "whisper" }, { "id": 10, "seek": 5100, "start": 68.0, "end": 76.0, "text": "那這裡面首先我們調用了配音或者說音色的這個模型,是謙問的TTS的一個開源模型。", "tokens": [ 51215, 4184, 18771, 8833, 36490, 5884, 32943, 9254, 2289, 38846, 18034, 31148, 4622, 18034, 17673, 1546, 6287, 41908, 39823, 11, 1541, 4802, 247, 11361, 1546, 51, 7327, 1546, 8990, 8949, 47402, 41908, 39823, 1543, 51615 ], "temperature": 0, "avg_logprob": -0.11461719306739601, "compression_ratio": 1.4353312302839116, "no_speech_prob": 1.8985370567325965e-11, "source": "whisper" }, { "id": 11, "seek": 5100, "start": 76.0, "end": 80.0, "text": "大家可以在本地的算力上進行運行,也是免費的。", "tokens": [ 51615, 6868, 6723, 3581, 8802, 10928, 1546, 19497, 13486, 5708, 18214, 8082, 32772, 8082, 11, 22021, 2347, 235, 50071, 1546, 1543, 51815 ], "temperature": 0, "avg_logprob": -0.11461719306739601, "compression_ratio": 1.4353312302839116, "no_speech_prob": 1.8985370567325965e-11, "source": "whisper" }, { "id": 12, "seek": 8000, "start": 80.0, "end": 85.0, "text": "那除了之外呢也是支持了你自己用自己的原聲進行克隆。", "tokens": [ 50365, 4184, 32999, 2289, 9574, 12022, 6240, 22021, 35488, 2289, 2166, 17645, 9254, 17645, 1546, 19683, 29454, 18214, 8082, 24881, 10673, 228, 1543, 50615 ], "temperature": 0, "avg_logprob": -0.16275492774115669, "compression_ratio": 1.3427230046948357, "no_speech_prob": 1.6465969213119713e-11, "source": "whisper" }, { "id": 13, "seek": 8000, "start": 85.0, "end": 88.0, "text": "直接用你自己的聲音然後去發出相關聲音來。", "tokens": [ 50615, 43297, 9254, 2166, 17645, 1546, 29454, 18034, 10213, 6734, 14637, 7781, 15106, 14899, 29454, 18034, 3763, 1543, 50765 ], "temperature": 0, "avg_logprob": -0.16275492774115669, "compression_ratio": 1.3427230046948357, "no_speech_prob": 1.6465969213119713e-11, "source": "whisper" }, { "id": 14, "seek": 8000, "start": 88.0, "end": 93.0, "text": "這裡的標準音色比如你是個中文的人,那你說英文那你的標準音色。", "tokens": [ 50765, 18771, 1546, 46299, 28220, 18034, 17673, 36757, 32526, 3338, 5975, 17174, 29979, 11, 4184, 42920, 27869, 17174, 4184, 18961, 46299, 28220, 18034, 17673, 1543, 51015 ], "temperature": 0, "avg_logprob": -0.16275492774115669, "compression_ratio": 1.3427230046948357, "no_speech_prob": 1.6465969213119713e-11, "source": "whisper" }, { "id": 15, "seek": 8000, "start": 93.0, "end": 97.0, "text": "配下來沒有我們這種標準音色做的那麼流暢吧。", "tokens": [ 51015, 38846, 27769, 6963, 5884, 33005, 46299, 28220, 18034, 17673, 10907, 1546, 20504, 27854, 23826, 95, 6062, 1543, 51215 ], "temperature": 0, "avg_logprob": -0.16275492774115669, "compression_ratio": 1.3427230046948357, "no_speech_prob": 1.6465969213119713e-11, "source": "whisper" }, { "id": 16, "seek": 9700, "start": 97.0, "end": 102.0, "text": "所以說這裡會有兩個版本,一個是標準音色版本,一個是克隆原聲的一個版本。", "tokens": [ 50365, 7239, 4622, 18771, 6236, 2412, 34623, 42096, 8802, 11, 8990, 1541, 46299, 28220, 18034, 17673, 42096, 8802, 11, 8990, 1541, 24881, 10673, 228, 19683, 29454, 1546, 8990, 42096, 8802, 1543, 50615 ], "temperature": 0, "avg_logprob": -0.15305890916269035, "compression_ratio": 1.3351063829787233, "no_speech_prob": 1.381118656940128e-11, "source": "whisper" }, { "id": 17, "seek": 9700, "start": 102.0, "end": 108.0, "text": "當然還有一個GPT的Soviet,用這個Soviet也可以去,有一些語音也是可以的。", "tokens": [ 50615, 21707, 15569, 8990, 38, 47, 51, 1546, 50, 5179, 302, 11, 9254, 6287, 50, 5179, 302, 47935, 6734, 11, 32241, 13824, 31348, 18034, 22021, 6723, 1546, 1543, 50915 ], "temperature": 0, "avg_logprob": -0.15305890916269035, "compression_ratio": 1.3351063829787233, "no_speech_prob": 1.381118656940128e-11, "source": "whisper", "text_original": "當然還有一個GPT的Sovet,用這個Sovet也可以去,有一些語音也是可以的。", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "id": 18, "seek": 9700, "start": 108.0, "end": 112.0, "text": "但是這裡的那個語音與音色裡面就會更少一些。", "tokens": [ 50915, 11189, 18771, 1546, 20754, 31348, 18034, 34127, 17673, 32399, 47778, 19002, 15686, 38515, 1543, 51115 ], "temperature": 0, "avg_logprob": -0.15305890916269035, "compression_ratio": 1.3351063829787233, "no_speech_prob": 1.381118656940128e-11, "source": "whisper", "text_original": "但是這裡的那個語音與色裡面就會更少一些。", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "id": 19, "seek": 11200, "start": 112.0, "end": 119.0, "text": "那我這裡推薦大家用這個QN3TTS,如果你要用自己的語音,那你就用QN3TTS克隆的原聲。", "tokens": [ 50365, 4184, 1654, 18771, 33597, 27746, 99, 6868, 9254, 6287, 48, 45, 18, 51, 7327, 11, 13119, 32085, 9254, 17645, 1546, 31348, 18034, 11, 4184, 41045, 9254, 48, 45, 18, 51, 7327, 24881, 10673, 228, 1546, 19683, 29454, 1543, 50715 ], "temperature": 0, "avg_logprob": -0.16028164372299658, "compression_ratio": 1.2780269058295963, "no_speech_prob": 1.7731862853032077e-11, "source": "whisper" }, { "id": 20, "seek": 11200, "start": 119.0, "end": 123.0, "text": "好這是第一個關於配音和音色的一個選擇。", "tokens": [ 50715, 2131, 16600, 40760, 14899, 19488, 38846, 18034, 12565, 18034, 17673, 1546, 8990, 21625, 13167, 229, 1543, 50915 ], "temperature": 0, "avg_logprob": -0.16028164372299658, "compression_ratio": 1.2780269058295963, "no_speech_prob": 1.7731862853032077e-11, "source": "whisper" }, { "id": 21, "seek": 11200, "start": 123.0, "end": 129.0, "text": "那此外呢我們還可以選擇字幕的一個顏色。比如我這裡選了中音的字幕,或者說音中的字幕。", "tokens": [ 50915, 4184, 17947, 12022, 6240, 5884, 7824, 6723, 21625, 13167, 229, 22381, 40211, 1546, 8990, 10178, 237, 17673, 1543, 36757, 1654, 18771, 21625, 2289, 5975, 18034, 1546, 22381, 40211, 11, 31148, 4622, 18034, 5975, 1546, 22381, 40211, 1543, 51215 ], "temperature": 0, "avg_logprob": -0.16028164372299658, "compression_ratio": 1.2780269058295963, "no_speech_prob": 1.7731862853032077e-11, "source": "whisper" }, { "id": 22, "seek": 12900, "start": 129.0, "end": 136.0, "text": "中音的字幕就是說中文在上面,英文在下面,那反過來就是英文在上面和中文在下面。", "tokens": [ 50365, 5975, 18034, 1546, 22381, 40211, 26222, 5975, 17174, 3581, 49750, 11, 27869, 17174, 3581, 47150, 11, 4184, 22138, 49950, 5620, 27869, 17174, 3581, 49750, 12565, 5975, 17174, 3581, 47150, 1543, 50715 ], "temperature": 0, "avg_logprob": -0.11528317133585612, "compression_ratio": 1.6357615894039734, "no_speech_prob": 1.4442971124628556e-11, "source": "whisper" }, { "id": 23, "seek": 12900, "start": 136.0, "end": 143.0, "text": "那也支持其他語種,包括什麼德語、法語、日語、繁體中文都是沒有問題。", "tokens": [ 50715, 4184, 6404, 35488, 9572, 5000, 31348, 16516, 11, 41828, 7598, 35898, 31348, 1231, 11148, 31348, 1231, 6890, 31348, 1231, 23141, 223, 23987, 5975, 17174, 22796, 6963, 17197, 1543, 51065 ], "temperature": 0, "avg_logprob": -0.11528317133585612, "compression_ratio": 1.6357615894039734, "no_speech_prob": 1.4442971124628556e-11, "source": "whisper" }, { "id": 24, "seek": 12900, "start": 143.0, "end": 150.0, "text": "那字幕的顏色大家也可以看到,比如我們這裡有中文和英文字幕的一個顏色,我們可以在這裡進行選擇。", "tokens": [ 51065, 4184, 22381, 40211, 1546, 10178, 237, 17673, 6868, 47935, 18032, 11, 36757, 5884, 18771, 2412, 5975, 17174, 12565, 27869, 17174, 22381, 40211, 1546, 8990, 10178, 237, 17673, 11, 5884, 6723, 3581, 18771, 18214, 8082, 21625, 13167, 229, 1543, 51415 ], "temperature": 0, "avg_logprob": -0.11528317133585612, "compression_ratio": 1.6357615894039734, "no_speech_prob": 1.4442971124628556e-11, "source": "whisper" }, { "id": 25, "seek": 12900, "start": 150.0, "end": 153.0, "text": "比如你中文要一個顏色,英文要一個顏色也是可以的。", "tokens": [ 51415, 36757, 2166, 5975, 17174, 4275, 8990, 10178, 237, 17673, 11, 27869, 17174, 4275, 8990, 10178, 237, 17673, 22021, 6723, 1546, 1543, 51565 ], "temperature": 0, "avg_logprob": -0.11528317133585612, "compression_ratio": 1.6357615894039734, "no_speech_prob": 1.4442971124628556e-11, "source": "whisper" }, { "id": 26, "seek": 12900, "start": 153.0, "end": 159.0, "text": "好,講到字幕的顏色,除此之外呢還有這個叫電影級的一個潤色。", "tokens": [ 51565, 2131, 11, 11932, 4511, 22381, 40211, 1546, 10178, 237, 17673, 11, 32999, 17947, 9574, 12022, 6240, 15569, 6287, 19855, 20545, 16820, 26020, 1546, 8990, 35622, 97, 17673, 1543, 51865 ], "temperature": 0, "avg_logprob": -0.11528317133585612, "compression_ratio": 1.6357615894039734, "no_speech_prob": 1.4442971124628556e-11, "source": "whisper", "text_original": "好,講到字幕的顏色,除此之外呢還有這個叫電影集的一個潤色。", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "id": 27, "seek": 15900, "start": 159.0, "end": 166.0, "text": "可以把你的字幕做一些深度的AI校對,以及做相關的一個潤色處理,會表現更好一些。", "tokens": [ 50365, 6723, 16075, 18961, 22381, 40211, 10907, 38515, 24043, 13127, 1546, 48698, 49684, 2855, 11, 40282, 10907, 15106, 14899, 1546, 8990, 35622, 97, 17673, 39289, 13876, 11, 6236, 17571, 9581, 19002, 2131, 38515, 1543, 50715 ], "temperature": 0, "avg_logprob": -0.13950764886264144, "compression_ratio": 1.4115853658536586, "no_speech_prob": 1.8095412321339488e-11, "source": "whisper" }, { "id": 28, "seek": 15900, "start": 166.0, "end": 174.0, "text": "這個也是推薦大家去勾上的。但這個勾上的話,它調用的就是我們API站內的大語言模型。", "tokens": [ 50715, 6287, 22021, 33597, 27746, 99, 6868, 6734, 7978, 122, 5708, 1546, 1543, 8395, 6287, 7978, 122, 5708, 21358, 11, 11284, 32943, 9254, 1546, 5620, 5884, 4715, 40, 34155, 28472, 1546, 3582, 31348, 12009, 41908, 39823, 1543, 51115 ], "temperature": 0, "avg_logprob": -0.13950764886264144, "compression_ratio": 1.4115853658536586, "no_speech_prob": 1.8095412321339488e-11, "source": "whisper" }, { "id": 29, "seek": 15900, "start": 174.0, "end": 178.0, "text": "我們可以看一下這裡有一個語音識別和翻譯功能。", "tokens": [ 51115, 5884, 6723, 28324, 18771, 2412, 8990, 31348, 18034, 43143, 16158, 12565, 42716, 11823, 107, 32311, 8225, 1543, 51315 ], "temperature": 0, "avg_logprob": -0.13950764886264144, "compression_ratio": 1.4115853658536586, "no_speech_prob": 1.8095412321339488e-11, "source": "whisper" }, { "id": 30, "seek": 15900, "start": 178.0, "end": 182.0, "text": "那翻譯的話其實是需要用到我們的一個API的。", "tokens": [ 51315, 4184, 42716, 11823, 107, 21358, 14139, 1541, 35748, 9254, 4511, 33641, 8990, 4715, 40, 1546, 1543, 51515 ], "temperature": 0, "avg_logprob": -0.13950764886264144, "compression_ratio": 1.4115853658536586, "no_speech_prob": 1.8095412321339488e-11, "source": "whisper" }, { "id": 31, "seek": 15900, "start": 182.0, "end": 188.0, "text": "那我們也可以直接進到我們的API分發站裡面來,然後選擇自己的一個令牌,點擊控制台。", "tokens": [ 51515, 46714, 47935, 43297, 18214, 4511, 33641, 4715, 40, 6627, 14637, 34155, 32399, 3763, 11, 10213, 21625, 13167, 229, 17645, 1546, 8990, 49061, 6935, 234, 11, 8216, 37813, 48707, 25491, 15433, 1543, 51815 ], "temperature": 0, "avg_logprob": -0.13950764886264144, "compression_ratio": 1.4115853658536586, "no_speech_prob": 1.8095412321339488e-11, "source": "whisper" }, { "id": 32, "seek": 18800, "start": 188.0, "end": 192.0, "text": "然後API密鑰。我們複製自己的這個令牌。", "tokens": [ 50365, 10213, 4715, 40, 47838, 4275, 1543, 5884, 164, 97, 229, 8083, 121, 17645, 1546, 6287, 49061, 6935, 234, 1543, 50565 ], "temperature": 0, "avg_logprob": -0.17540512084960938, "compression_ratio": 1.125, "no_speech_prob": 1.891149910282497e-11, "source": "whisper", "text_original": "然後API密要。我們複製自己的這個令牌。", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "id": 33, "seek": 18800, "start": 192.0, "end": 197.0, "text": "然後填入到這個框框裡面來。填入之後我們點擊保存令牌。", "tokens": [ 50565, 10213, 13331, 104, 14028, 4511, 6287, 21416, 228, 21416, 228, 32399, 3763, 1543, 13331, 104, 14028, 26362, 5884, 8216, 37813, 24302, 39781, 49061, 6935, 234, 1543, 50815 ], "temperature": 0, "avg_logprob": -0.17540512084960938, "compression_ratio": 1.125, "no_speech_prob": 1.891149910282497e-11, "source": "whisper" }, { "id": 34, "seek": 18800, "start": 197.0, "end": 198.0, "text": "然後獲取模型。", "tokens": [ 50815, 10213, 29057, 110, 29436, 41908, 39823, 1543, 50865 ], "temperature": 0, "avg_logprob": -0.17540512084960938, "compression_ratio": 1.125, "no_speech_prob": 1.891149910282497e-11, "source": "whisper" }, { "id": 35, "seek": 19800, "start": 198.0, "end": 204.0, "text": "比如我在這裡選一個5.6的terra。", "tokens": [ 50365, 4184, 6287, 14010, 6240, 47778, 2412, 5884, 4715, 40, 3581, 28472, 1546, 38714, 41908, 39823, 1543, 36757, 40528, 18771, 21625, 8990, 20, 13, 21, 1546, 391, 424, 1543, 50665 ], "temperature": 0, "avg_logprob": -0.1673655546912851, "compression_ratio": 1.3717105263157894, "no_speech_prob": 1.7792725626186723e-11, "source": "whisper", "text_original": "那這個時候呢就會有我們API在內的全部模型。比如我在這裡選一個5.6的terra。", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "id": 36, "seek": 19800, "start": 204.0, "end": 208.0, "text": "或者說選一個luna。用來跑翻譯還是很方便的嘛。", "tokens": [ 50665, 31148, 4622, 21625, 8990, 75, 5051, 1543, 9254, 3763, 32585, 42716, 11823, 107, 25583, 4563, 9249, 27364, 1546, 20722, 1543, 50865 ], "temperature": 0, "avg_logprob": -0.1673655546912851, "compression_ratio": 1.3717105263157894, "no_speech_prob": 1.7792725626186723e-11, "source": "whisper" }, { "id": 37, "seek": 19800, "start": 208.0, "end": 214.0, "text": "那我就選這個terra了。然後OK,我們現在就已經全部準備好了。我們視頻也已經上傳了。對吧。", "tokens": [ 50865, 4184, 22020, 21625, 6287, 391, 424, 2289, 1543, 10213, 9443, 11, 5884, 12648, 3111, 25338, 38714, 36241, 12621, 1543, 5884, 27333, 29940, 6404, 25338, 5708, 40852, 2289, 1543, 47905, 1543, 51165 ], "temperature": 0, "avg_logprob": -0.1673655546912851, "compression_ratio": 1.3717105263157894, "no_speech_prob": 1.7792725626186723e-11, "source": "whisper" }, { "id": 38, "seek": 19800, "start": 214.0, "end": 220.0, "text": "或者說你直接通過鏈接去下載的。也可以。你直接鏈接下載之後,你那個視頻也會自動的加載過來。", "tokens": [ 51165, 31148, 4622, 2166, 43297, 19550, 8816, 46241, 230, 14468, 6734, 4438, 12290, 231, 1546, 1543, 47935, 1543, 2166, 43297, 46241, 230, 14468, 4438, 12290, 231, 26362, 11, 2166, 20754, 27333, 29940, 6404, 6236, 9722, 12572, 1546, 9990, 12290, 231, 49950, 1543, 51465 ], "temperature": 0, "avg_logprob": -0.1673655546912851, "compression_ratio": 1.3717105263157894, "no_speech_prob": 1.7792725626186723e-11, "source": "whisper" }, { "id": 39, "seek": 22000, "start": 220.0, "end": 227.0, "text": "就不需要你二次上傳了。好,那我們處理完成之後,我們只要點一下開始處理這個按鈕就可以。", "tokens": [ 50365, 3111, 1960, 35748, 2166, 11217, 9487, 5708, 40852, 2289, 1543, 2131, 11, 46714, 39289, 13876, 41509, 26362, 11, 5884, 14003, 4275, 8216, 8861, 21017, 39289, 13876, 6287, 26613, 41597, 243, 36781, 1543, 50715 ], "temperature": 0, "avg_logprob": -0.1299949259985061, "compression_ratio": 1.2836538461538463, "no_speech_prob": 1.927145769353711e-11, "source": "whisper" }, { "id": 40, "seek": 22000, "start": 227.0, "end": 236.0, "text": "我們點擊開始處理。那這裡會自動的進行配音環境的一個校對。如果你缺相關的環境,它也會自動的幫你去裝。", "tokens": [ 50715, 5884, 8216, 37813, 21017, 39289, 13876, 1543, 4184, 18771, 6236, 9722, 12572, 1546, 18214, 8082, 38846, 18034, 40063, 108, 47885, 1546, 8990, 49684, 2855, 1543, 45669, 38109, 118, 15106, 14899, 1546, 40063, 108, 47885, 11, 11284, 6404, 6236, 9722, 12572, 1546, 32187, 2166, 6734, 44453, 1543, 51165 ], "temperature": 0, "avg_logprob": -0.1299949259985061, "compression_ratio": 1.2836538461538463, "no_speech_prob": 1.927145769353711e-11, "source": "whisper" }, { "id": 41, "seek": 23600, "start": 236.0, "end": 244.0, "text": "那這個其實這個後戶端,我已經幫大家裝好了那個配音的環境。然後還有千問3的這個模型也給大家配好了。", "tokens": [ 50365, 4184, 6287, 14139, 6287, 5661, 1486, 114, 11957, 107, 11, 1654, 25338, 32187, 6868, 44453, 12621, 20754, 38846, 18034, 1546, 40063, 108, 47885, 1543, 10213, 15569, 20787, 11361, 18, 1546, 6287, 41908, 39823, 6404, 17798, 6868, 38846, 12621, 1543, 50765 ], "temperature": 0, "avg_logprob": -0.17573654992239815, "compression_ratio": 1.525993883792049, "no_speech_prob": 1.720824004125543e-11, "source": "whisper" }, { "id": 42, "seek": 23600, "start": 244.0, "end": 251.0, "text": "還有這個語音識別的whisper的模型。這裡一點需要大家需要注意。如果你本地有顯卡的話,你可以選這個median。", "tokens": [ 50765, 15569, 6287, 31348, 18034, 43143, 16158, 1546, 1363, 271, 610, 1546, 41908, 39823, 1543, 18771, 19202, 35748, 6868, 35748, 40794, 1543, 45669, 8802, 10928, 2412, 46605, 32681, 21358, 11, 42766, 21625, 6287, 1912, 952, 1543, 51115 ], "temperature": 0, "avg_logprob": -0.17573654992239815, "compression_ratio": 1.525993883792049, "no_speech_prob": 1.720824004125543e-11, "source": "whisper" }, { "id": 43, "seek": 23600, "start": 251.0, "end": 256.0, "text": "如果你本地沒有顯卡的話,你可以選這個small或者說tiny去做語音的識別。", "tokens": [ 51115, 45669, 8802, 10928, 6963, 46605, 32681, 21358, 11, 42766, 21625, 6287, 10817, 336, 31148, 4622, 83, 3519, 6734, 10907, 31348, 18034, 1546, 43143, 16158, 1543, 51365 ], "temperature": 0, "avg_logprob": -0.17573654992239815, "compression_ratio": 1.525993883792049, "no_speech_prob": 1.720824004125543e-11, "source": "whisper" }, { "id": 44, "seek": 23600, "start": 256.0, "end": 264.0, "text": "這個是需要本地算力的,包括上面這個千問3這個TTS本質都是算力的。那我可以給大家看一下。", "tokens": [ 51365, 41427, 35748, 8802, 10928, 19497, 13486, 1546, 11, 41828, 49750, 6287, 20787, 11361, 18, 6287, 51, 7327, 8802, 43450, 22796, 19497, 13486, 1546, 1543, 4184, 1654, 6723, 17798, 6868, 28324, 1543, 51765 ], "temperature": 0, "avg_logprob": -0.17573654992239815, "compression_ratio": 1.525993883792049, "no_speech_prob": 1.720824004125543e-11, "source": "whisper" }, { "id": 45, "seek": 26400, "start": 264.0, "end": 272.0, "text": "像我任務管理器裡面的一些東西。任務管理器點擊性能。然後大家可以看到這裡的英偉達顯卡是會直接去占用的。", "tokens": [ 50365, 12760, 1654, 26443, 39517, 23131, 13876, 34386, 32399, 1546, 38515, 26978, 1543, 26443, 39517, 23131, 13876, 34386, 8216, 37813, 21686, 8225, 1543, 10213, 6868, 6723, 18032, 18771, 1546, 27869, 7437, 231, 33086, 46605, 32681, 1541, 6236, 43297, 6734, 5322, 254, 9254, 1546, 1543, 50765 ], "temperature": 0, "avg_logprob": -0.15698323184496735, "compression_ratio": 1.3767705382436262, "no_speech_prob": 1.757861564644081e-11, "source": "whisper" }, { "id": 46, "seek": 26400, "start": 272.0, "end": 283.0, "text": "特別是macOS如果你是8GB的內存,你就不要想了。這個你基本跑不了。", "tokens": [ 50765, 45669, 8802, 10928, 6963, 46605, 32681, 11, 31148, 4622, 32526, 37065, 4367, 1546, 39035, 20854, 11, 4184, 2166, 18771, 6723, 42061, 8949, 7322, 19202, 1543, 38446, 1541, 37065, 4367, 13119, 32526, 23, 8769, 1546, 28472, 39781, 11, 41045, 11962, 7093, 2289, 1543, 6287, 2166, 37946, 32585, 47225, 1543, 51315 ], "temperature": 0, "avg_logprob": -0.15698323184496735, "compression_ratio": 1.3767705382436262, "no_speech_prob": 1.757861564644081e-11, "source": "whisper", "text_original": "如果你本地沒有顯卡,或者說你是macOS的設備,那你這裡可以把它開小一點。特別是macOS如果你是8GB的內存,你就不要想了。這個你基本跑不了。", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "id": 47, "seek": 26400, "start": 283.0, "end": 294.0, "text": "你至少也要有個16G或者說24G去運行,會比較流暢一些。而這個整個客戶端會比較大一點。大家需要注意,它大概是有10G", "tokens": [ 51315, 2166, 20844, 15686, 6404, 4275, 2412, 3338, 6866, 38, 31148, 4622, 7911, 38, 6734, 32772, 8082, 11, 6236, 25174, 27854, 23826, 95, 38515, 1543, 11070, 6287, 27662, 3338, 32316, 1486, 114, 11957, 107, 6236, 25174, 3582, 19202, 1543, 6868, 35748, 40794, 11, 11284, 32044, 1541, 2412, 3279, 38, 51865 ], "temperature": 0, "avg_logprob": -0.15698323184496735, "compression_ratio": 1.3767705382436262, "no_speech_prob": 1.757861564644081e-11, "source": "whisper" }, { "id": 48, "seek": 29400, "start": 294.0, "end": 302.0, "text": "一個Gb左右吧。大概這麼大。因為裡面有一些模型。然後還有就是除了說模型,還有就是一些依賴環境。", "tokens": [ 50365, 8990, 38, 65, 29457, 6062, 1543, 32044, 21269, 3582, 1543, 11471, 32399, 32241, 13824, 41908, 39823, 1543, 10213, 15569, 5620, 32999, 2289, 4622, 41908, 39823, 11, 15569, 5620, 38515, 3254, 251, 164, 7062, 40063, 108, 47885, 1543, 50765 ], "temperature": 0, "avg_logprob": -0.20108332778468277, "compression_ratio": 1.275, "no_speech_prob": 2.1447900747095794e-11, "source": "whisper" }, { "id": 49, "seek": 29400, "start": 302.0, "end": 312.0, "text": "Python的一些依賴環境。包括我們的配音模型和語音識別的這兩個模型。本質都在本地跑。你的這個雲端API只是用來跑那個翻譯的。", "tokens": [ 50765, 47, 88, 11943, 1546, 38515, 3254, 251, 164, 7062, 40063, 108, 47885, 1543, 41828, 33641, 38846, 18034, 41908, 39823, 12565, 31348, 18034, 43143, 16158, 1546, 2664, 34623, 41908, 39823, 1543, 8802, 43450, 7182, 3581, 8802, 10928, 32585, 1543, 18961, 6287, 6306, 110, 11957, 107, 4715, 40, 36859, 9254, 3763, 32585, 20754, 42716, 11823, 107, 1546, 1543, 51265 ], "temperature": 0, "avg_logprob": -0.20108332778468277, "compression_ratio": 1.275, "no_speech_prob": 2.1447900747095794e-11, "source": "whisper" }, { "id": 50, "seek": 31200, "start": 312.0, "end": 322.0, "text": "我覺得接一個API去跑翻譯,效果會更好一些。等這個流程跑完。我們去看一下它的效果就好。", "tokens": [ 50365, 17689, 14468, 8990, 4715, 40, 6734, 32585, 42716, 11823, 107, 11, 43076, 9319, 6236, 19002, 2131, 38515, 1543, 10187, 6287, 27854, 29649, 32585, 14128, 1543, 5884, 6734, 28324, 45224, 43076, 9319, 3111, 2131, 1543, 50865 ], "temperature": 0, "avg_logprob": -0.2028883197639562, "compression_ratio": 1.3088235294117647, "no_speech_prob": 1.752912918984162e-11, "source": "whisper" }, { "id": 51, "seek": 31200, "start": 322.0, "end": 332.0, "text": "好,我們來看一下這個任務快完成了。首先它做的是這個音頻的抽取。抽取完了音頻之後,進行語音的一個識別。", "tokens": [ 50865, 2131, 11, 5884, 3763, 28324, 6287, 26443, 39517, 10251, 41509, 2289, 1543, 36490, 11284, 10907, 24620, 6287, 18034, 29940, 1546, 46022, 29436, 1543, 46022, 29436, 41665, 18034, 29940, 26362, 11, 18214, 8082, 31348, 18034, 1546, 8990, 43143, 16158, 1543, 51365 ], "temperature": 0, "avg_logprob": -0.2028883197639562, "compression_ratio": 1.3088235294117647, "no_speech_prob": 1.752912918984162e-11, "source": "whisper" }, { "id": 52, "seek": 33200, "start": 332.0, "end": 339.0, "text": "然後整理了我們的一個字幕。然後再將字幕做了一個翻譯。因為我們的目標語言是英文。", "tokens": [ 50365, 10213, 27662, 13876, 2289, 33641, 8990, 22381, 40211, 1543, 10213, 8623, 34168, 22381, 40211, 10907, 2289, 8990, 42716, 11823, 107, 1543, 11471, 33641, 11386, 46299, 31348, 12009, 1541, 27869, 17174, 1543, 50715 ], "temperature": 0, "avg_logprob": -0.09621690354257259, "compression_ratio": 1.4958333333333333, "no_speech_prob": 1.7233268631566823e-11, "source": "whisper" }, { "id": 53, "seek": 33200, "start": 339.0, "end": 348.0, "text": "翻譯成了英語之後,再進行一個TTS的一個配音。配音完成之後,最後就是合成我們的音頻。", "tokens": [ 50715, 42716, 11823, 107, 11336, 2289, 27869, 31348, 26362, 11, 8623, 18214, 8082, 8990, 51, 7327, 1546, 8990, 38846, 18034, 1543, 38846, 18034, 41509, 26362, 11, 20578, 5620, 14245, 11336, 33641, 18034, 29940, 1543, 51165 ], "temperature": 0, "avg_logprob": -0.09621690354257259, "compression_ratio": 1.4958333333333333, "no_speech_prob": 1.7233268631566823e-11, "source": "whisper" }, { "id": 54, "seek": 33200, "start": 348.0, "end": 356.0, "text": "那這次已經進行到了最後這一個片段,就是分成了70個片段嘛。那這裡就是具體的一個工作流程。", "tokens": [ 51165, 4184, 2664, 9487, 25338, 18214, 8082, 21381, 20578, 2664, 8990, 16668, 28427, 11, 5620, 6627, 11336, 2289, 5867, 3338, 16668, 28427, 20722, 1543, 4184, 18771, 5620, 39806, 23987, 1546, 8990, 41315, 27854, 29649, 1543, 51565 ], "temperature": 0, "avg_logprob": -0.09621690354257259, "compression_ratio": 1.4958333333333333, "no_speech_prob": 1.7233268631566823e-11, "source": "whisper" }, { "id": 55, "seek": 35600, "start": 356.0, "end": 365.0, "text": "整個操作其實就很簡單。你上傳的視頻,然後選好你的目標語種。對吧。還有一些字幕的顏色。或者說電影集的一些潤色。", "tokens": [ 50365, 27662, 3338, 13167, 235, 11914, 14139, 3111, 4563, 31995, 1543, 2166, 5708, 40852, 1546, 27333, 29940, 11, 10213, 21625, 2131, 18961, 11386, 46299, 31348, 16516, 1543, 47905, 1543, 15569, 38515, 22381, 40211, 1546, 10178, 237, 17673, 1543, 31148, 4622, 20545, 16820, 26020, 1546, 38515, 35622, 97, 17673, 1543, 50815 ], "temperature": 0, "avg_logprob": -0.15899326719086745, "compression_ratio": 1.436950146627566, "no_speech_prob": 2.1389561996598694e-11, "source": "whisper" }, { "id": 56, "seek": 35600, "start": 365.0, "end": 374.0, "text": "然後配一個那個翻譯的模型。那就在這裡。我這裡選了一個翻譯的模型。最後就是點一下開始處理。", "tokens": [ 50815, 10213, 38846, 8990, 20754, 42716, 11823, 107, 1546, 41908, 39823, 1543, 46887, 3581, 18771, 1543, 1654, 18771, 21625, 2289, 8990, 42716, 11823, 107, 1546, 41908, 39823, 1543, 20578, 5620, 8216, 8861, 21017, 39289, 13876, 1543, 51265 ], "temperature": 0, "avg_logprob": -0.15899326719086745, "compression_ratio": 1.436950146627566, "no_speech_prob": 2.1389561996598694e-11, "source": "whisper" }, { "id": 57, "seek": 35600, "start": 374.0, "end": 380.0, "text": "然後這裡就會進行自動化的處理。現在再進行混合音軌,並且封裝到這個視頻裡面來。", "tokens": [ 51265, 10213, 18771, 47778, 18214, 8082, 9722, 12572, 23756, 1546, 39289, 13876, 1543, 12648, 8623, 18214, 8082, 48640, 14245, 18034, 14454, 234, 11, 35934, 20334, 1530, 223, 44453, 4511, 6287, 27333, 29940, 32399, 3763, 1543, 51565 ], "temperature": 0, "avg_logprob": -0.15899326719086745, "compression_ratio": 1.436950146627566, "no_speech_prob": 2.1389561996598694e-11, "source": "whisper" }, { "id": 58, "seek": 35600, "start": 380.0, "end": 386.0, "text": "那我們可以點開去看一下。到時候應該就會在我們這一個Subflow的,", "tokens": [ 51565, 46714, 6723, 8216, 8949, 6734, 28324, 1543, 4511, 14010, 26087, 47778, 3581, 5884, 2664, 8990, 39582, 10565, 1546, 11, 51865 ], "temperature": 0, "avg_logprob": -0.15899326719086745, "compression_ratio": 1.436950146627566, "no_speech_prob": 2.1389561996598694e-11, "source": "whisper" }, { "id": 59, "seek": 38600, "start": 386.0, "end": 392.0, "text": "這個項目裡面。那這裡會生成一個視頻。這裡是剛才的這一個視頻。", "tokens": [ 50365, 6287, 8313, 227, 11386, 32399, 1543, 4184, 18771, 6236, 8244, 11336, 8990, 27333, 29940, 1543, 18771, 1541, 16940, 18888, 1546, 2664, 8990, 27333, 29940, 1543, 50665 ], "temperature": 0, "avg_logprob": -0.17248370912339953, "compression_ratio": 1.658878504672897, "no_speech_prob": 2.5597109501451776e-11, "source": "whisper" }, { "id": 60, "seek": 38600, "start": 392.0, "end": 395.0, "text": "這是原版的視頻,我們可以先打開來看一眼。", "tokens": [ 50665, 16600, 19683, 42096, 1546, 27333, 29940, 11, 5884, 6723, 10108, 12467, 8949, 3763, 4200, 2257, 25281, 1543, 50815 ], "temperature": 0, "avg_logprob": -0.17248370912339953, "compression_ratio": 1.658878504672897, "no_speech_prob": 2.5597109501451776e-11, "source": "whisper" }, { "id": 61, "seek": 38600, "start": 395.0, "end": 398.0, "text": "然後通過這個。", "tokens": [ 50815, 10213, 19550, 8816, 6287, 1543, 50965 ], "temperature": 0, "avg_logprob": -0.17248370912339953, "compression_ratio": 1.658878504672897, "no_speech_prob": 2.5597109501451776e-11, "source": "whisper" }, { "id": 62, "seek": 38600, "start": 398.0, "end": 402.0, "text": "這是原版的這個視頻。", "tokens": [ 50965, 16600, 19683, 42096, 1546, 6287, 27333, 29940, 1543, 51165 ], "temperature": 0, "avg_logprob": -0.17248370912339953, "compression_ratio": 1.658878504672897, "no_speech_prob": 2.5597109501451776e-11, "source": "whisper" }, { "id": 63, "seek": 38600, "start": 402.0, "end": 408.0, "text": "然後它是完全沒有字幕的。是我之前自己錄的一個視頻。", "tokens": [ 51165, 10213, 11284, 1541, 37100, 6963, 22381, 40211, 1546, 1543, 42614, 32442, 17645, 9567, 226, 1546, 8990, 27333, 29940, 1543, 51465 ], "temperature": 0, "avg_logprob": -0.17248370912339953, "compression_ratio": 1.658878504672897, "no_speech_prob": 2.5597109501451776e-11, "source": "whisper" }, { "id": 64, "seek": 38600, "start": 408.0, "end": 413.0, "text": "對吧。然後完全沒有字幕。那這裡生成的是這個音中版本的。", "tokens": [ 51465, 47905, 1543, 10213, 37100, 6963, 22381, 40211, 1543, 4184, 18771, 8244, 11336, 24620, 6287, 18034, 5975, 42096, 8802, 1546, 1543, 51715 ], "temperature": 0, "avg_logprob": -0.17248370912339953, "compression_ratio": 1.658878504672897, "no_speech_prob": 2.5597109501451776e-11, "source": "whisper" }, { "id": 65, "seek": 41300, "start": 413.0, "end": 417.0, "text": "那這裡還沒有。這裡已經生成完成了。我們可以看一下。", "tokens": [ 50365, 4184, 18771, 7824, 6963, 1543, 18771, 25338, 8244, 11336, 41509, 2289, 1543, 5884, 6723, 28324, 1543, 50565 ], "temperature": 0, "avg_logprob": -0.3162647027235765, "compression_ratio": 0.9782608695652174, "no_speech_prob": 2.7677396832737067e-11, "source": "whisper" }, { "id": 66, "seek": 41300, "start": 417.0, "end": 419.0, "text": "打開這個。", "tokens": [ 50565, 12467, 8949, 6287, 1543, 50665 ], "temperature": 0, "avg_logprob": -0.3162647027235765, "compression_ratio": 0.9782608695652174, "no_speech_prob": 2.7677396832737067e-11, "source": "whisper" }, { "id": 67, "seek": 41900, "start": 419.0, "end": 421.0, "text": "打開這個。", "tokens": [ 50365, 12467, 8949, 6287, 1543, 50465 ], "temperature": 0, "avg_logprob": -0.35323783329554964, "compression_ratio": 1.0, "no_speech_prob": 2.51515336335828e-11, "source": "whisper" }, { "id": 68, "seek": 41900, "start": 421.0, "end": 425.0, "text": "音中字幕的一個版本。那下面兩個是它的那個字幕。", "tokens": [ 50465, 18034, 5975, 22381, 40211, 1546, 8990, 42096, 8802, 1543, 4184, 47150, 34623, 1541, 45224, 20754, 22381, 40211, 1543, 50665 ], "temperature": 0, "avg_logprob": -0.35323783329554964, "compression_ratio": 1.0, "no_speech_prob": 2.51515336335828e-11, "source": "whisper" }, { "id": 69, "seek": 42500, "start": 425.0, "end": 432.0, "text": "SRT的一個字幕。那我們打開這個視頻。這個視頻已經燒錄好了。就是把字幕那些東西都加進去了。", "tokens": [ 50365, 50, 49, 51, 1546, 8990, 22381, 40211, 1543, 46714, 12467, 8949, 6287, 27333, 29940, 1543, 6287, 27333, 29940, 25338, 24184, 240, 9567, 226, 12621, 1543, 5620, 16075, 22381, 40211, 4184, 13824, 26978, 7182, 9990, 18214, 45190, 1543, 50715 ], "temperature": 0, "avg_logprob": -0.18899582843391263, "compression_ratio": 1.5, "no_speech_prob": 2.6699120345141658e-11, "source": "whisper" }, { "id": 70, "seek": 42500, "start": 432.0, "end": 437.0, "text": "大家可以看到。這裡不只是說。", "tokens": [ 50715, 6868, 6723, 18032, 1543, 18771, 1960, 36859, 4622, 1543, 50965 ], "temperature": 0, "avg_logprob": -0.18899582843391263, "compression_ratio": 1.5, "no_speech_prob": 2.6699120345141658e-11, "source": "whisper" }, { "id": 71, "seek": 42500, "start": 437.0, "end": 441.0, "text": "我們把這個中文的視頻換成了一個英文的視頻。", "tokens": [ 50965, 5884, 16075, 6287, 5975, 17174, 1546, 27333, 29940, 36338, 11336, 2289, 8990, 27869, 17174, 1546, 27333, 29940, 1543, 51165 ], "temperature": 0, "avg_logprob": -0.18899582843391263, "compression_ratio": 1.5, "no_speech_prob": 2.6699120345141658e-11, "source": "whisper" }, { "id": 72, "seek": 42500, "start": 441.0, "end": 445.0, "text": "而且配上了英文加中文的一個字幕。那這裡的音色。我用的是。", "tokens": [ 51165, 22942, 38846, 5708, 2289, 27869, 17174, 9990, 5975, 17174, 1546, 8990, 22381, 40211, 1543, 4184, 18771, 1546, 18034, 17673, 1543, 1654, 9254, 24620, 1543, 51365 ], "temperature": 0, "avg_logprob": -0.18899582843391263, "compression_ratio": 1.5, "no_speech_prob": 2.6699120345141658e-11, "source": "whisper" }, { "id": 73, "seek": 44500, "start": 445.0, "end": 453.0, "text": "千文的TTS的一個音色。然後選的是第一個男聲的一個聲音。OK。那這是。", "tokens": [ 50365, 20787, 17174, 1546, 51, 7327, 1546, 8990, 18034, 17673, 1543, 10213, 21625, 24620, 40760, 25243, 29454, 1546, 8990, 29454, 18034, 1543, 9443, 1543, 4184, 16600, 1543, 50765 ], "temperature": 0, "avg_logprob": -0.21353866577148437, "compression_ratio": 1.308108108108108, "no_speech_prob": 2.5373772527537142e-11, "source": "whisper" }, { "id": 74, "seek": 44500, "start": 453.0, "end": 458.0, "text": "這一個演示。那大家可以看一下。就每一塊。", "tokens": [ 50765, 2664, 8990, 31382, 25696, 1543, 4184, 6868, 6723, 28324, 1543, 3111, 23664, 2257, 26268, 1543, 51015 ], "temperature": 0, "avg_logprob": -0.21353866577148437, "compression_ratio": 1.308108108108108, "no_speech_prob": 2.5373772527537142e-11, "source": "whisper" }, { "id": 75, "seek": 44500, "start": 458.0, "end": 461.0, "text": "大家都配。配上了相關的音中字幕。", "tokens": [ 51015, 46525, 38846, 1543, 38846, 5708, 2289, 15106, 14899, 1546, 18034, 5975, 22381, 40211, 1543, 51165 ], "temperature": 0, "avg_logprob": -0.21353866577148437, "compression_ratio": 1.308108108108108, "no_speech_prob": 2.5373772527537142e-11, "source": "whisper" }, { "id": 76, "seek": 44500, "start": 461.0, "end": 464.0, "text": "大家一般用的比較多的還是。", "tokens": [ 51165, 6868, 2257, 49640, 9254, 1546, 25174, 6392, 1546, 25583, 1543, 51315 ], "temperature": 0, "avg_logprob": -0.21353866577148437, "compression_ratio": 1.308108108108108, "no_speech_prob": 2.5373772527537142e-11, "source": "whisper" }, { "id": 77, "seek": 46400, "start": 464.0, "end": 467.0, "text": "國內小夥伴。一般用的比較多的就是這個中音字幕。", "tokens": [ 50365, 8053, 28472, 7322, 1787, 98, 7384, 112, 1543, 2257, 49640, 9254, 1546, 25174, 6392, 1546, 5620, 6287, 5975, 18034, 22381, 40211, 1543, 50515 ], "temperature": 0, "avg_logprob": -0.15651840263313346, "compression_ratio": 1.4664536741214058, "no_speech_prob": 2.2173927624602463e-11, "source": "whisper" }, { "id": 78, "seek": 46400, "start": 467.0, "end": 472.0, "text": "就是中音雙語字幕。然後目標語種是中文。那他就不需要這個配音。", "tokens": [ 50515, 5620, 5975, 18034, 6306, 247, 31348, 22381, 40211, 1543, 10213, 11386, 46299, 31348, 16516, 1541, 5975, 17174, 1543, 4184, 5000, 3111, 1960, 35748, 6287, 38846, 18034, 1543, 50765 ], "temperature": 0, "avg_logprob": -0.15651840263313346, "compression_ratio": 1.4664536741214058, "no_speech_prob": 2.2173927624602463e-11, "source": "whisper" }, { "id": 79, "seek": 46400, "start": 472.0, "end": 475.0, "text": "和這個選音色。因為就是用你自己的一個聲音。", "tokens": [ 50765, 12565, 6287, 21625, 18034, 17673, 1543, 11471, 5620, 9254, 2166, 17645, 1546, 8990, 29454, 18034, 1543, 50915 ], "temperature": 0, "avg_logprob": -0.15651840263313346, "compression_ratio": 1.4664536741214058, "no_speech_prob": 2.2173927624602463e-11, "source": "whisper" }, { "id": 80, "seek": 46400, "start": 475.0, "end": 479.0, "text": "那只有你選擇目標語種的時候。他才會出來配音。", "tokens": [ 50915, 4184, 35244, 2166, 21625, 13167, 229, 11386, 46299, 31348, 16516, 20643, 1543, 5000, 18888, 6236, 29741, 38846, 18034, 1543, 51115 ], "temperature": 0, "avg_logprob": -0.15651840263313346, "compression_ratio": 1.4664536741214058, "no_speech_prob": 2.2173927624602463e-11, "source": "whisper" }, { "id": 81, "seek": 46400, "start": 479.0, "end": 482.0, "text": "OK。這是一塊。第二塊就是說。", "tokens": [ 51115, 9443, 1543, 16600, 2257, 26268, 1543, 19693, 26268, 26222, 1543, 51265 ], "temperature": 0, "avg_logprob": -0.15651840263313346, "compression_ratio": 1.4664536741214058, "no_speech_prob": 2.2173927624602463e-11, "source": "whisper" }, { "id": 82, "seek": 46400, "start": 482.0, "end": 486.0, "text": "當大家在做這個subflow。特別是。", "tokens": [ 51265, 13118, 6868, 3581, 10907, 6287, 30131, 10565, 1543, 38446, 1541, 1543, 51465 ], "temperature": 0, "avg_logprob": -0.15651840263313346, "compression_ratio": 1.4664536741214058, "no_speech_prob": 2.2173927624602463e-11, "source": "whisper" }, { "id": 83, "seek": 46400, "start": 486.0, "end": 491.0, "text": "在進行這個視頻下的時候。你的那個視頻的那個cookie可能會到期。", "tokens": [ 51465, 3581, 18214, 8082, 6287, 27333, 29940, 4438, 20643, 1543, 18961, 20754, 27333, 29940, 1546, 20754, 66, 39611, 16657, 6236, 4511, 16786, 1543, 51715 ], "temperature": 0, "avg_logprob": -0.15651840263313346, "compression_ratio": 1.4664536741214058, "no_speech_prob": 2.2173927624602463e-11, "source": "whisper" }, { "id": 84, "seek": 49100, "start": 491.0, "end": 494.0, "text": "你需要更換自己的cookie。那怎麼去更換呢?", "tokens": [ 50365, 2166, 35748, 19002, 36338, 17645, 1546, 66, 39611, 1543, 4184, 11614, 6734, 19002, 36338, 6240, 30, 50515 ], "temperature": 0, "avg_logprob": -0.1512540266817848, "compression_ratio": 1.5034013605442176, "no_speech_prob": 1.601907322068552e-11, "source": "whisper" }, { "id": 85, "seek": 49100, "start": 494.0, "end": 498.0, "text": "是需要在這個cookies的項目文件夾內有兩個文件。", "tokens": [ 50515, 1541, 35748, 3581, 6287, 66, 1212, 530, 1546, 8313, 227, 11386, 17174, 20485, 1787, 122, 28472, 2412, 34623, 17174, 20485, 1543, 50715 ], "temperature": 0, "avg_logprob": -0.1512540266817848, "compression_ratio": 1.5034013605442176, "no_speech_prob": 1.601907322068552e-11, "source": "whisper" }, { "id": 86, "seek": 49100, "start": 498.0, "end": 501.0, "text": "一個是youtube的cookie。一個是bilibili的cookies。", "tokens": [ 50715, 8990, 1541, 88, 346, 1977, 1546, 66, 39611, 1543, 8990, 1541, 65, 12722, 2312, 1546, 66, 1212, 530, 1543, 50865 ], "temperature": 0, "avg_logprob": -0.1512540266817848, "compression_ratio": 1.5034013605442176, "no_speech_prob": 1.601907322068552e-11, "source": "whisper" }, { "id": 87, "seek": 49100, "start": 501.0, "end": 504.0, "text": "那我們怎麼去獲取這個cookie呢?", "tokens": [ 50865, 46714, 11614, 6734, 29057, 110, 29436, 6287, 66, 39611, 6240, 30, 51015 ], "temperature": 0, "avg_logprob": -0.1512540266817848, "compression_ratio": 1.5034013605442176, "no_speech_prob": 1.601907322068552e-11, "source": "whisper" }, { "id": 88, "seek": 49100, "start": 504.0, "end": 509.0, "text": "首先第一步我們是需要在那個我們的導航頁。然後往下滑。", "tokens": [ 51015, 36490, 18049, 31429, 5884, 1541, 35748, 3581, 20754, 33641, 25647, 10256, 103, 8313, 223, 1543, 10213, 29510, 4438, 15868, 239, 1543, 51265 ], "temperature": 0, "avg_logprob": -0.1512540266817848, "compression_ratio": 1.5034013605442176, "no_speech_prob": 1.601907322068552e-11, "source": "whisper" }, { "id": 89, "seek": 49100, "start": 509.0, "end": 514.0, "text": "往下找。找到有一個叫chrome插件的商店。當然你用的edg。", "tokens": [ 51265, 29510, 4438, 25085, 1543, 25085, 4511, 2412, 8990, 19855, 339, 11505, 11673, 240, 20485, 1546, 45581, 22469, 1543, 21707, 2166, 9254, 1546, 292, 70, 1543, 51515 ], "temperature": 0, "avg_logprob": -0.1512540266817848, "compression_ratio": 1.5034013605442176, "no_speech_prob": 1.601907322068552e-11, "source": "whisper" }, { "id": 90, "seek": 49100, "start": 514.0, "end": 518.0, "text": "你可以用這個edg的插件商店。然後進到這裡面來之後。", "tokens": [ 51515, 42766, 9254, 6287, 292, 70, 1546, 11673, 240, 20485, 45581, 22469, 1543, 10213, 18214, 4511, 18771, 8833, 3763, 26362, 1543, 51715 ], "temperature": 0, "avg_logprob": -0.1512540266817848, "compression_ratio": 1.5034013605442176, "no_speech_prob": 1.601907322068552e-11, "source": "whisper" }, { "id": 91, "seek": 51800, "start": 518.0, "end": 523.0, "text": "我們搜索getcookies。搜索getcookies。大家可以看到我安裝的是這個版本。", "tokens": [ 50365, 5884, 162, 14987, 7732, 95, 847, 66, 1212, 530, 1543, 162, 14987, 7732, 95, 847, 66, 1212, 530, 1543, 6868, 6723, 18032, 1654, 16206, 44453, 24620, 6287, 42096, 8802, 1543, 50615 ], "temperature": 0, "avg_logprob": -0.12179814444647895, "compression_ratio": 1.4822695035460993, "no_speech_prob": 2.176460400793445e-11, "source": "whisper" }, { "id": 92, "seek": 51800, "start": 523.0, "end": 528.0, "text": "叫getcookies.textclean。這個東西。那怎麼去用呢?", "tokens": [ 50615, 19855, 847, 66, 1212, 530, 13, 25111, 2160, 282, 1543, 6287, 26978, 1543, 4184, 11614, 6734, 9254, 6240, 30, 50865 ], "temperature": 0, "avg_logprob": -0.12179814444647895, "compression_ratio": 1.4822695035460993, "no_speech_prob": 2.176460400793445e-11, "source": "whisper" }, { "id": 93, "seek": 51800, "start": 528.0, "end": 533.0, "text": "比如我這裡是youtube。那我只要在這個位置點一下。", "tokens": [ 50865, 36757, 1654, 18771, 1541, 88, 346, 1977, 1543, 4184, 1654, 14003, 4275, 3581, 6287, 11160, 34719, 8216, 8861, 1543, 51115 ], "temperature": 0, "avg_logprob": -0.12179814444647895, "compression_ratio": 1.4822695035460993, "no_speech_prob": 2.176460400793445e-11, "source": "whisper" }, { "id": 94, "seek": 51800, "start": 533.0, "end": 536.0, "text": "然後我們就可以copy他的youtube的cookies。", "tokens": [ 51115, 10213, 5884, 36781, 13084, 88, 31309, 88, 346, 1977, 1546, 66, 1212, 530, 1543, 51265 ], "temperature": 0, "avg_logprob": -0.12179814444647895, "compression_ratio": 1.4822695035460993, "no_speech_prob": 2.176460400793445e-11, "source": "whisper" }, { "id": 95, "seek": 51800, "start": 536.0, "end": 541.0, "text": "那我們把這個cookies複製一下。然後填到這一個文檔裡面來。", "tokens": [ 51265, 46714, 16075, 6287, 66, 1212, 530, 164, 97, 229, 8083, 121, 8861, 1543, 10213, 13331, 104, 4511, 2664, 8990, 17174, 40741, 242, 32399, 3763, 1543, 51515 ], "temperature": 0, "avg_logprob": -0.12179814444647895, "compression_ratio": 1.4822695035460993, "no_speech_prob": 2.176460400793445e-11, "source": "whisper" }, { "id": 96, "seek": 51800, "start": 541.0, "end": 545.0, "text": "然後雙擊打開。那填到這裡面來。把原來這個刪除。", "tokens": [ 51515, 10213, 6306, 247, 37813, 12467, 8949, 1543, 4184, 13331, 104, 4511, 18771, 8833, 3763, 1543, 16075, 19683, 3763, 6287, 2437, 103, 32999, 1543, 51715 ], "temperature": 0, "avg_logprob": -0.12179814444647895, "compression_ratio": 1.4822695035460993, "no_speech_prob": 2.176460400793445e-11, "source": "whisper" }, { "id": 97, "seek": 54500, "start": 545.0, "end": 549.0, "text": "如果你沒有的話。那你只貼進來。然後保存上。", "tokens": [ 50365, 45669, 6963, 21358, 1543, 4184, 2166, 14003, 11561, 120, 18214, 3763, 1543, 10213, 24302, 39781, 5708, 1543, 50565 ], "temperature": 0, "avg_logprob": -0.17576333407698005, "compression_ratio": 1.4605263157894737, "no_speech_prob": 1.7612921537901727e-11, "source": "whisper" }, { "id": 98, "seek": 54500, "start": 549.0, "end": 551.0, "text": "這個時候我們的cookies就更換成功了。", "tokens": [ 50565, 6287, 14010, 33641, 66, 1212, 530, 3111, 19002, 36338, 11336, 32311, 2289, 1543, 50665 ], "temperature": 0, "avg_logprob": -0.17576333407698005, "compression_ratio": 1.4605263157894737, "no_speech_prob": 1.7612921537901727e-11, "source": "whisper" }, { "id": 99, "seek": 54500, "start": 551.0, "end": 555.0, "text": "YouTube。Bilibili是一樣的。你只要隨機打開Bilibili的網頁。", "tokens": [ 50665, 3223, 41173, 1543, 33, 12722, 2312, 1541, 22977, 1546, 1543, 2166, 14003, 4275, 48890, 17543, 12467, 8949, 33, 12722, 2312, 1546, 26352, 8313, 223, 1543, 50865 ], "temperature": 0, "avg_logprob": -0.17576333407698005, "compression_ratio": 1.4605263157894737, "no_speech_prob": 1.7612921537901727e-11, "source": "whisper" }, { "id": 100, "seek": 54500, "start": 555.0, "end": 559.0, "text": "然後用同樣的方法。比如我這裡隨便打開一個Bilibili的網站。", "tokens": [ 50865, 10213, 9254, 13089, 6180, 1546, 9249, 11148, 1543, 36757, 1654, 18771, 48890, 27364, 12467, 8949, 8990, 33, 12722, 2312, 1546, 26352, 34155, 1543, 51065 ], "temperature": 0, "avg_logprob": -0.17576333407698005, "compression_ratio": 1.4605263157894737, "no_speech_prob": 1.7612921537901727e-11, "source": "whisper" }, { "id": 101, "seek": 54500, "start": 559.0, "end": 563.0, "text": "然後我只要點一下這個按鈕。然後點擊copy。", "tokens": [ 51065, 10213, 1654, 14003, 4275, 8216, 8861, 6287, 26613, 41597, 243, 1543, 10213, 8216, 37813, 13084, 88, 1543, 51265 ], "temperature": 0, "avg_logprob": -0.17576333407698005, "compression_ratio": 1.4605263157894737, "no_speech_prob": 1.7612921537901727e-11, "source": "whisper" }, { "id": 102, "seek": 54500, "start": 563.0, "end": 570.0, "text": "就可以把b站的cookies給他copy下來。然後這樣的話就可以再線去把雲端的視頻給他下載下來。", "tokens": [ 51265, 36781, 16075, 65, 34155, 1546, 66, 1212, 530, 17798, 5000, 13084, 88, 27769, 1543, 10213, 8377, 21358, 36781, 8623, 31699, 6734, 16075, 6306, 110, 11957, 107, 1546, 27333, 29940, 17798, 5000, 4438, 12290, 231, 27769, 1543, 51615 ], "temperature": 0, "avg_logprob": -0.17576333407698005, "compression_ratio": 1.4605263157894737, "no_speech_prob": 1.7612921537901727e-11, "source": "whisper" }, { "id": 103, "seek": 57000, "start": 570.0, "end": 574.0, "text": "當然如果你是本地視頻你直接上傳就好了。", "tokens": [ 50365, 21707, 13119, 32526, 8802, 10928, 27333, 29940, 2166, 43297, 5708, 40852, 3111, 12621, 1543, 50565 ], "temperature": 0, "avg_logprob": -0.1801903247833252, "compression_ratio": 1.2851985559566788, "no_speech_prob": 1.8225585970976788e-11, "source": "whisper" }, { "id": 104, "seek": 57000, "start": 574.0, "end": 581.0, "text": "有關於我們這一款叫Subflow語幕的這一款。", "tokens": [ 50565, 9443, 1543, 2664, 5620, 8802, 16786, 25246, 30855, 17798, 6868, 1546, 2412, 14899, 19488, 5884, 32260, 48798, 19855, 39582, 10565, 34127, 44, 1816, 1546, 32260, 48798, 1543, 50915 ], "temperature": 0, "avg_logprob": -0.1801903247833252, "compression_ratio": 1.2851985559566788, "no_speech_prob": 1.8225585970976788e-11, "source": "whisper", "text_original": "OK。這就是本期我想分享給大家的有關於我們這一款叫Subflow與Mood的這一款。", "correction_applied": true, "correction_source": "llm_correct_a_presegmentation" }, { "id": 105, "seek": 57000, "start": 581.0, "end": 587.0, "text": "直接提供了我們Windows和MacOS的一個客戶端。大家直接下載客戶端去使用就好了。", "tokens": [ 50915, 43297, 20949, 3254, 249, 2289, 5884, 45813, 1509, 12565, 34355, 4367, 1546, 8990, 32316, 1486, 114, 11957, 107, 1543, 6868, 43297, 4438, 12290, 231, 32316, 1486, 114, 11957, 107, 6734, 22982, 9254, 3111, 12621, 1543, 51215 ], "temperature": 0, "avg_logprob": -0.1801903247833252, "compression_ratio": 1.2851985559566788, "no_speech_prob": 1.8225585970976788e-11, "source": "whisper" }, { "id": 106, "seek": 57000, "start": 587.0, "end": 593.0, "text": "好。大家對這個項目感興趣的話可以幫我們點一點小信心。點一下fork。", "tokens": [ 51215, 2131, 1543, 6868, 2855, 6287, 8313, 227, 11386, 9709, 44089, 39835, 21358, 6723, 32187, 5884, 8216, 19202, 7322, 17665, 7945, 1543, 8216, 8861, 69, 1284, 1543, 51515 ], "temperature": 0, "avg_logprob": -0.1801903247833252, "compression_ratio": 1.2851985559566788, "no_speech_prob": 1.8225585970976788e-11, "source": "whisper" }, { "id": 107, "seek": 59300, "start": 593.0, "end": 597.0, "text": "謝謝。這就是本期視頻的全部內容。謝謝大家。", "tokens": [ 50365, 11003, 1543, 2664, 5620, 8802, 16786, 27333, 29940, 1546, 38714, 28472, 25750, 1543, 11003, 6868, 1543, 50565 ], "temperature": 0, "avg_logprob": -0.14284495303505346, "compression_ratio": 0.9130434782608695, "no_speech_prob": 1.427759473149326e-11, "source": "whisper" } ], "transcript_selection": { "schema": "videonote.transcript_selection.v1", "status": "SELECTED", "requested_language": "zh", "duration": 597.472688, "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": 108, "multilingual": false, "first_start": 0.0, "final_end": 597.0, "duration_coverage": 0.999209, "maximum_gap_seconds": 0.0, "inverted_count": 0, "language": { "requested": "zh", "matched": true, "latin": 405, "han": 2799, "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": 597.473, "rejected_segment_count": 0, "rejected_segments": [] } }, "presegmentation_correction": { "applied": true, "correction_count": 10, "deterministic_fix_count": 0 } }