planeta ai logo
Darmowy szablon automatyzacji

Przekształć artykuły BBC News w podcasty za pomocą Hugging Face i Google Gemini

3066
27 dni temu
22
bloków


Przekształć artykuły BBC News w podcasty za pomocą Hugging Face i Google Gemini

Zautomatyzuj proces tworzenia podcastów na podstawie artykułów BBC News dzięki temu gotowemu szablonowi n8n. Idealne rozwiązanie dla osób, które chcą szybko generować treści audio bez konieczności posiadania zaawansowanych umiejętności technicznych.

Dla kogo jest ten szablon?

Ten szablon jest idealny dla:

  • Twórców treści, którzy chcą szybko produkować podcasty na podstawie aktualnych wydarzeń.
  • Studentów poszukujących efektywnego sposobu na tworzenie materiałów audio do projektów lub zadań.
  • Osób zainteresowanych generowaniem własnych podcastów bez konieczności posiadania wiedzy technicznej.

Informacje o konfiguracji

Wymagane kroki do uruchomienia:

  1. Zainstaluj n8n: Jeśli jeszcze tego nie zrobiłeś, pobierz i zainstaluj n8n ze strony n8n.io.
  2. Zaimportuj workflow: Skopiuj kod JSON tego workflow i zaimportuj go do swojej instancji n8n.
  3. Skonfiguruj dane uwierzytelniające:
    • Gemini API: Skonfiguruj swoje dane uwierzytelniające API Gemini w węzłach LLM workflow.
    • Token Hugging Face: Uzyskaj token dostępu z Hugging Face i dodaj go do węzła HTTP Request dla modelu tekst-na-mowę.
  4. Dostosuj (opcjonalnie):
    • Kryteria filtrowania: Dostosuj węzeł News Classifier, aby precyzyjniej dobierać artykuły zgodnie z Twoimi preferencjami.
    • Opcje wyjściowe: Zmodyfikuj workflow, aby zapisywać wygenerowany plik audio w usłudze przechowywania w chmurze lub publikować go na platformie hostingowej podcastów.

Wymagania wstępne

  • Aktywna instancja n8n.
  • Podstawowa znajomość workflow n8n (nie wymaga kodowania).
  • Dane uwierzytelniające API Gemini oraz konto Hugging Face z tokenem dostępu.

Jakie problemy rozwiązuje?

Ten workflow eliminuje konieczność ręcznego tworzenia podcastów z artykułów newsowych. Automatyzuje cały proces, od pobierania i filtrowania wiadomości po generowanie finalnego pliku audio.

Jakie są korzyści?

  • Oszczędność czasu: Twórz podcasty w minutach, a nie godzinach.
  • Łatwość użycia: Nie wymaga umiejętności kodowania ani technicznych.
  • Możliwość dostosowania: Adaptuj workflow do swoich specyficznych potrzeb i preferencji.
  • Ekonomiczność: Wykorzystaj darmowe lub niskokosztowe usługi takie jak Gemini i Hugging Face.

Jak to działa?

  1. Workflow pobiera artykuły newsowe ze strony BBC.
  2. Filtruje artykuły pod kątem ich przydatności do podcastu.
  3. Ekstrahuje pełną treść wybranych artykułów.
  4. Wykorzystuje Gemini LLM do stworzenia skryptu podcastu.
  5. Konwertuje skrypt na mowę za pomocą modelu tekst-na-mowę Hugging Face.
  6. Finalny plik audio podcastu jest gotowy do użycia.

Węzły w workflow

  • Fetch BBC News Page: Pobiera główną stronę BBC News.
  • News Classifier: Kategoryzuje artykuły newsowe przy użyciu Gemini LLM.
  • Fetch BBC News Detail: Ekstrahuje szczegółową treść z odpowiednich artykułów.
  • Basic Podcast LLM Chain: Generuje skrypt podcastu przy użyciu Gemini LLM.
  • HTTP Request: Konwertuje skrypt na mowę przy użyciu Hugging Face.

Przykłady zastosowań

Ten szablon automatyzacji może być wykorzystany w wielu różnych scenariuszach. Oto kilka potencjalnych zastosowań:

  • Tworzenie codziennych podcastów informacyjnych na podstawie najnowszych wiadomości.
  • Generowanie materiałów edukacyjnych dla studentów w formie audio.
  • Automatyczne przygotowywanie podcastów tematycznych (np. technologicznych, sportowych).
  • Tworzenie treści audio dla osób niedowidzących lub preferujących słuchanie zamiast czytania.
  • Produkcja podcastów firmowych na podstawie branżowych newsów.
  • Generowanie materiałów do nauki języków obcych na podstawie aktualnych wydarzeń.
  • Tworzenie podcastów podróżniczych na podstawie newsów z różnych regionów świata.

Dodatkowe wskazówki

  • Zapoznaj się z dokumentacją n8n i zasobami społeczności, aby poznać bardziej zaawansowane opcje dostosowywania.
  • Eksperymentuj z różnymi kryteriami filtrowania i promptami LLM, aby osiągnąć pożądany styl podcastu.


   Skopiuj kod szablonu   
{"meta":{"instanceId":"5287ddd2fa569cf8e4c5a724666246a45305c032a19bb677c9e4b963d365f84b","templateCredsSetupCompleted":true},"nodes":[{"id":"95c798a4-bc34-4219-b7c3-6b4a4070886b","name":"When clicking ‘Test workflow’","type":"n8n-nodes-base.manualTrigger","position":[-320,1080],"parameters":{},"typeVersion":1},{"id":"09987590-1ec2-48d4-aa04-32b85addd9e6","name":"Split Out","type":"n8n-nodes-base.splitOut","position":[420,1080],"parameters":{"options":{},"fieldToSplitOut":"newsTitle"},"typeVersion":1},{"id":"758e3f60-01dc-46c7-bb53-7460eaed92e3","name":"Extract News Block","type":"n8n-nodes-base.html","position":[220,1080],"parameters":{"options":{},"operation":"extractHtmlContent","extractionValues":{"values":[{"key":"newsTitle","cssSelector":".eGcloy","returnArray":true,"returnValue":"html"}]}},"typeVersion":1.2},{"id":"20440f9a-a40c-4419-af6d-383de041d078","name":"Extract News Content","type":"n8n-nodes-base.html","position":[600,1080],"parameters":{"options":{},"operation":"extractHtmlContent","dataPropertyName":"newsTitle","extractionValues":{"values":[{"key":"title","cssSelector":"h2"},{"key":"link","attribute":"href","cssSelector":"a","returnValue":"attribute"},{"key":"description","cssSelector":".kYtujW"}]}},"typeVersion":1.2},{"id":"47a50ada-127a-4037-8fe7-41c0caebb3de","name":"Aggregate","type":"n8n-nodes-base.aggregate","position":[2000,1060],"parameters":{"options":{},"aggregate":"aggregateAllItemData"},"typeVersion":1},{"id":"33050a43-842d-464d-b227-a6c2c870c0af","name":"Fetch BBC News Detail","type":"n8n-nodes-base.httpRequest","position":[1400,1060],"parameters":{"url":"=https://www.bbc.com{{ $json.link }}","options":{}},"typeVersion":4.2},{"id":"646bfe6b-cac6-4177-9b59-dc205b44b7eb","name":"Extract Detail","type":"n8n-nodes-base.html","position":[1600,1060],"parameters":{"options":{},"operation":"extractHtmlContent","extractionValues":{"values":[{"key":"newsDetail","cssSelector":".dlWCEZ .fYAfXe","returnArray":true}]}},"typeVersion":1.2},{"id":"6e14f528-1e94-411f-8601-3c713d492aa9","name":"Filter Empty Detail","type":"n8n-nodes-base.filter","position":[1800,1060],"parameters":{"options":{},"conditions":{"options":{"version":2,"leftValue":"","caseSensitive":true,"typeValidation":"strict"},"combinator":"and","conditions":[{"id":"7066e88c-03da-4196-b1c5-80bc16fa3fc6","operator":{"type":"array","operation":"notEmpty","singleValue":true},"leftValue":"={{ $json.newsDetail }}","rightValue":""}]}},"typeVersion":2.2},{"id":"5863e420-2392-468a-8e03-5d4c273168e0","name":"If script exists","type":"n8n-nodes-base.if","position":[2620,1060],"parameters":{"options":{},"conditions":{"options":{"version":1,"leftValue":"","caseSensitive":true,"typeValidation":"strict"},"combinator":"and","conditions":[{"id":"2e968b41-88f7-4b28-9837-af50ae130979","operator":{"type":"string","operation":"exists","singleValue":true},"leftValue":"=voice_id {{ $json.output.podcast_script }}","rightValue":""}]}},"typeVersion":2},{"id":"90b370d3-5712-401d-b769-490014e2b17c","name":"Basic Podcast LLM Chain","type":"@n8n/n8n-nodes-langchain.chainLlm","position":[2200,1060],"parameters":{"text":"=News Articles:{{ $json.data.map(item => item.newsDetail) }}","messages":{"messageValues":[{"message":"= n*Convert the following one or multiple news articles into a podcast script formatted for direct use in ElevenLabs. If there is only one news piece, transform it into a compelling and engaging narrative. If multiple news stories are provided, structure them like a news bulletin, presenting each piece sequentially with smooth transitions. Avoid a formal or dry tone; instead, use a natural, conversational, and warm style. The podcast should feel dynamic, engaging, and informative while maintaining a storytelling approach.* nn- *Ensure the script is formatted as a single, continuous text block suitable for direct speech synthesis.* n- *Start with an engaging introduction that sets the tone for the podcast.* n- *Narrate each news story smoothly, with natural transitions between segments.* n- *End with a closing statement that leaves the listener informed and engaged.* n- *Output must be in JSON format, with the full script as a single string under the key `"podcast_script"`.* nn---nn### **Input Format:** n```jsonn{n "news_articles": [n {n "title": "First News Title",n "content": "First news article content..."n },n {n "title": "Second News Title",n "content": "Second news article content..."n }n ]n}n```nn---nnExpected JSON Output Format:n n{n "podcast_script": "Welcome to today's news podcast! We have some exciting stories lined up for you. Let's start with our first story. [First news article content rewritten in a conversational, engaging style]... Moving on to our next topic... [Second news article content rewritten dynamically]... That’s all for today’s news bulletin! Stay informed and see you next time."n}nnn "}]},"promptType":"define","hasOutputParser":true},"typeVersion":1.5},{"id":"24c212c2-6d06-4fe1-841b-bc52a21060b1","name":"Gemini","type":"@n8n/n8n-nodes-langchain.lmChatGoogleGemini","position":[1600,1600],"parameters":{"options":{},"modelName":"models/gemini-2.0-pro-exp-02-05"},"credentials":{"googlePalmApi":{"id":"5x46RlCURyTUmbGW","name":"Google Gemini(PaLM) Api account 2"}},"typeVersion":1},{"id":"02e9f1ee-dc80-403c-8c19-0e6f918cf8ed","name":"Output Parser","type":"@n8n/n8n-nodes-langchain.outputParserStructured","position":[2360,1280],"parameters":{"jsonSchemaExample":"{nt"podcast_script": "California"n}"},"typeVersion":1.2},{"id":"395ddac7-b2a4-48c5-b2d3-d21078d29c54","name":"News Classifier","type":"@n8n/n8n-nodes-langchain.textClassifier","position":[980,1080],"parameters":{"options":{},"inputText":"=I will only send the headline as input:n{{ $json.title }} {{ $json.description }}","categories":{"categories":[{"category":"Suitable","description":"=Role: News Content Suitability Assessor (Positive)nnTask: Determine if the given news headline is highly likely to be suitable for storytelling, focusing on positive and engaging aspects.nnCriteria (Focus on what makes it suitable):nnCuriosity and Interest: Does the headline present an event, discovery, or information that is likely to pique the curiosity of a broad audience? Does it have a "wow" factor or relate to a significant trend?nnStorytelling Potential: Does the headline lend itself well to narrative expansion? Could it be the start of an engaging story, a key point in a developing situation, or a surprising conclusion?nnPositive or Neutral Tone: Is the headline generally positive, neutral, or focused on solutions/progress? (Avoid headlines that primarily focus on conflict, negativity, or routine events).nnRelevance: Does the headline touch upon topics that are relevant to a wide audience, such as health, science, technology, interesting discoveries, or positive global events?nnOutput Format:nnHeadline: [Original news headline]nnSuitable: [Yes / No] (Only say "Yes" if strongly confident)nnReason (Brief): [Briefly explain why it's likely suitable, focusing on the positive aspects.]nnExample (for the LLM to learn from):nnHeadline: "Scientists Discover New Species of Butterfly in Amazon Rainforest"nnSuitable: YesnnReason: Discovery, biodiversity, positive natural event, intriguing.nnHeadline: "Stock Market Experiences Minor Fluctuations"nnSuitable: NonnReason: Routine economic event, lacks general interest."},{"category":"Not Suitable","description":"=Role: News Content Filter (Negative)nnTask: Identify news headlines that are clearly unsuitable for storytelling due to negative content, lack of general interest, or ethical concerns.nnCriteria (Focus on what makes it unsuitable):nnNegative Content: Does the headline contain violence, crime, accidents, death, suffering, or other traumatic events?nnPolitical/Economic Routine: Does the headline focus on routine political announcements, standard economic reports (like minor market changes), or internal political disputes?nnDivisive or Harmful Content: Does the headline contain hate speech, discrimination, strong political bias, or potentially harmful misinformation?nnLack of General Interest: Is the headline highly niche, specific to a very small group, or about a topic unlikely to interest a broad audience?nnOutput Format:nnHeadline: [Original news headline]nnNot Suitable: [Yes / No] (Only say "Yes" if strongly confident)nnReason (Brief): [Briefly explain why it's clearly unsuitable.]nnExample (for the LLM to learn from):nnHeadline: "Local Politician Announces Campaign Platform"nnNot Suitable: YesnnReason: Routine political event, lacks broad appeal.nnHeadline: "Car Crash Results in Minor Injuries"nnNot Suitable: YesnnReason: Negative event (accident), though thankfully not severe."}]}},"typeVersion":1},{"id":"13fac9ed-688c-4af9-a810-d49a74b98c22","name":"Fetch BBC News Page","type":"n8n-nodes-base.httpRequest","position":[-60,1080],"parameters":{"url":"https://www.bbc.com/","options":{},"responseFormat":"string"},"typeVersion":1},{"id":"e2aa33f3-aa7c-4a9d-ac3c-32f9f5872606","name":"Sticky Note","type":"n8n-nodes-base.stickyNote","position":[-120,920],"parameters":{"width":500,"height":340,"content":"## This node fetches the main BBC News page, which contains links to various news articles."},"typeVersion":1},{"id":"0821b944-44cb-41ed-b5ff-70f99018c5dc","name":"Sticky Note1","type":"n8n-nodes-base.stickyNote","position":[960,840],"parameters":{"color":2,"width":340,"height":360,"content":"## This node uses a Gemini LLM to classify news articles based on their titles and descriptions. It determines if the content is suitable for a podcast.nn"},"typeVersion":1},{"id":"d32b2ebb-0a4d-4d27-9262-894ab7a65cce","name":"Sticky Note2","type":"n8n-nodes-base.stickyNote","position":[1360,820],"parameters":{"color":3,"width":400,"height":420,"content":"## This node fetches the detailed content of the news articles that were classified as suitable for a podcast."},"typeVersion":1},{"id":"e6e1d180-b2c2-4b62-a611-7c039037ed69","name":"Sticky Note3","type":"n8n-nodes-base.stickyNote","position":[2180,880],"parameters":{"color":4,"width":340,"height":320,"content":"## This node uses a Gemini LLM to convert the news articles into a podcast script.n"},"typeVersion":1},{"id":"d8776355-967d-4875-b948-25792f6f38ec","name":"Sticky Note4","type":"n8n-nodes-base.stickyNote","position":[2840,920],"parameters":{"color":5,"width":360,"height":300,"content":"## It structures the script for direct use with the Hugging Face text-to-speech model."},"typeVersion":1},{"id":"631a2caf-c640-41df-9215-2b542de51ccb","name":"Sticky Note5","type":"n8n-nodes-base.stickyNote","position":[-660,740],"parameters":{"width":460,"height":280,"content":"## 3rd Party Application Requirements:nn### Gemini nYou've already set up a Gemini LLM. No access token is needed for this.n### Hugging Facen You'll need an access token for the Hugging Face text-to-speech model n"},"typeVersion":1},{"id":"655e6799-5b7c-4747-b3a9-d01b47f5cba8","name":"Limit 10 Items","type":"n8n-nodes-base.limit","position":[800,1080],"parameters":{"maxItems":10},"typeVersion":1},{"id":"64d011d2-9c51-4f1f-a3b8-edf3fcbc6710","name":"Hugging Face Text-to-Speech.","type":"n8n-nodes-base.httpRequest","position":[2900,1060],"parameters":{"url":"https://router.huggingface.co/hf-inference/models/facebook/mms-tts-eng","method":"POST","options":{},"sendBody":true,"authentication":"predefinedCredentialType","bodyParameters":{"parameters":[{"name":"inputs","value":"={{ $json.output.podcast_script }}"}]},"nodeCredentialType":"huggingFaceApi"},"credentials":{"huggingFaceApi":{"id":"FF4PO5RYOJqZ0vhQ","name":"HuggingFaceApi account"}},"typeVersion":4.2}],"pinData":{},"connections":{"Gemini":{"ai_languageModel":[[{"node":"News Classifier","type":"ai_languageModel","index":0},{"node":"Basic Podcast LLM Chain","type":"ai_languageModel","index":0}]]},"Aggregate":{"main":[[{"node":"Basic Podcast LLM Chain","type":"main","index":0}]]},"Split Out":{"main":[[{"node":"Extract News Content","type":"main","index":0}]]},"Output Parser":{"ai_outputParser":[[{"node":"Basic Podcast LLM Chain","type":"ai_outputParser","index":0}]]},"Extract Detail":{"main":[[{"node":"Filter Empty Detail","type":"main","index":0}]]},"Limit 10 Items":{"main":[[{"node":"News Classifier","type":"main","index":0}]]},"News Classifier":{"main":[[{"node":"Fetch BBC News Detail","type":"main","index":0}]]},"If script exists":{"main":[[{"node":"Hugging Face Text-to-Speech.","type":"main","index":0}],[]]},"Extract News Block":{"main":[[{"node":"Split Out","type":"main","index":0}]]},"Fetch BBC News Page":{"main":[[{"node":"Extract News Block","type":"main","index":0}]]},"Filter Empty Detail":{"main":[[{"node":"Aggregate","type":"main","index":0}]]},"Extract News Content":{"main":[[{"node":"Limit 10 Items","type":"main","index":0}]]},"Fetch BBC News Detail":{"main":[[{"node":"Extract Detail","type":"main","index":0}]]},"Basic Podcast LLM Chain":{"main":[[{"node":"If script exists","type":"main","index":0}]]},"When clicking ‘Test workflow’":{"main":[[{"node":"Fetch BBC News Page","type":"main","index":0}]]}}}
  • API
  • Request
  • URL
  • Build
  • cURL
  • extract
  • template
  • table
  • LangChain
  • json
  • zod
Planeta AI 2025 
magic-wandmenu linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram