[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article-mlops-zoomcamp-path-to-production-ml-zh":3,"article-related-mlops-zoomcamp-path-to-production-ml-zh":32,"series-industry-1ca3cf77-7688-45c3-ad99-ecf7c0ec7f54":84},{"id":4,"slug":5,"title":6,"content":7,"summary":8,"source":9,"source_url":10,"author":11,"image_url":12,"cover_image":12,"category":13,"language":14,"translated_content":11,"related_article_id":15,"keywords":16,"key_takeaways":24,"views":28,"created_at":29,"published_at":30,"topic_cluster_id":31},"1ca3cf77-7688-45c3-ad99-ecf7c0ec7f54","mlops-zoomcamp-path-to-production-ml-zh","MLOps Zoomcamp 把模型帶上線的完整路線","\u003Cp data-speakable=\"summary\">這門免費課程用 9 個模組帶你把\u003Ca href=\"\u002Ftag\u002F機器學習\">機器學習\u003C\u002Fa>模型做成可部署、可監控的 production 系統。\u003C\u002Fp>\u003Cp>\u003Ca href=\"\u002Ftag\u002Fmlops\">MLOps\u003C\u002Fa> Zoomcamp 是一門免費的 9 週課程，\u003Ca href=\"\u002Ftag\u002Fgithub\">GitHub\u003C\u002Fa> 上有 14.8k stars，重點不是再教一次模型訓練，而是把你從實驗腳本一路帶到部署與監控。看完下面 5 項，你可以判斷自己缺的是模型管理、pipeline、部署方式，還是工程化習慣。\u003C\u002Fp>\u003Ctable>\u003Cthead>\u003Ctr>\u003Cth>項目\u003C\u002Fth>\u003Cth>重點\u003C\u002Fth>\u003Cth>可比規格\u003C\u002Fth>\u003C\u002Ftr>\u003C\u002Fthead>\u003Ctbody>\u003Ctr>\u003Ctd>課程長度\u003C\u002Ftd>\u003Ctd>學習節奏\u003C\u002Ftd>\u003Ctd>9 週\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>GitHub 熱度\u003C\u002Ftd>\u003Ctd>社群關注\u003C\u002Ftd>\u003Ctd>14.8k stars\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>模組數\u003C\u002Ftd>\u003Ctd>內容範圍\u003C\u002Ftd>\u003Ctd>9 個模組\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>專案收尾\u003C\u002Ftd>\u003Ctd>實作成果\u003C\u002Ftd>\u003Ctd>1 個 final project\u003C\u002Ftd>\u003C\u002Ftr>\u003C\u002Ftbody>\u003C\u002Ftable>\u003Ch2>1. 先把 MLOps 的全貌看懂\u003C\u002Fh2>\u003Cp>這門課先回答一個最關鍵的問題：模型訓練完之後，接下來到底要做什麼。開場模組會講 MLOps 的基本概念、課程結構，並用 \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FDataTalksClub\u002Fmlops-zoomcamp\" target=\"_blank\" rel=\"noopener noreferrer\">MLOps Zoomcamp\u003C\u002Fa> 的 NY Taxi 範例把整條流程串起來。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781542984202-6g6y.png\" alt=\"MLOps Zoomcamp 把模型帶上線的完整路線\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>如果你已經會基本機器學習，但\u003Ca href=\"\u002Fnews\u002Fclarity-act-reshaping-crypto-before-law-2026-zh\">還沒\u003C\u002Fa>碰過 production 流程，這一段最適合先看。它也明確設了門檻：Python、\u003Ca href=\"\u002Ftag\u002Fdocker\">Docker\u003C\u002Fa>、命令列工具，還有大約一年的程式經驗，代表這不是入門 ML 課，而是把既有技能接到實戰管線。\u003C\u002Fp>\u003Cul>\u003Cli>適合：資料科學家、ML 工程師、軟體工程師\u003C\u002Fli>\u003Cli>形式：預錄影片加作業\u003C\u002Fli>\u003Cli>主軸：MLOps maturity model、環境設定、課程導覽\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>2. 先管住實驗，才管得住模型\u003C\u002Fh2>\u003Cp>很多團隊卡住的第一步，不是訓練不出模型，而是忘了哪次實驗用了什麼參數。這一\u003Ca href=\"\u002Fnews\u002Fmetabot-8-module-stack-turns-metakpk-into-closed-loop-zh\">模組把\u003C\u002Fa> experiment tracking、\u003Ca href=\"https:\u002F\u002Fmlflow.org\u002F\" target=\"_blank\" rel=\"noopener noreferrer\">MLflow\u003C\u002Fa>、模型儲存與載入、model registry 放在一起，讓你能比較 run、保留版本、追蹤結果。\u003C\u002Fp>\u003Cp>如果你曾經在 notebook、腳本、雲端實驗之間迷路，這段會很有感。它的價值在於把一次性的訓練行為，\u003Ca href=\"\u002Fnews\u002Fmidjourney-v8-1-default-model-update-zh\">變成\u003C\u002Fa>可重現、可查詢、可交接的流程。\u003C\u002Fp>\u003Ccode>重點：experiment tracking \u002F MLflow basics \u002F model registry \u002F save-load \u002F homework\u003C\u002Fcode>\u003Ch2>3. 把零散腳本變成可跑的 pipeline\u003C\u002Fh2>\u003Cp>當實驗管理穩定後，課程接著談 orchestration，也就是怎麼讓資料處理、訓練、驗證按順序可靠執行。這一段的核心不是炫技，而是把分散的腳本整理成能重複跑的 ML pipeline。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781542993039-kkaf.png\" alt=\"MLOps Zoomcamp 把模型帶上線的完整路線\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>對需要減少人工操作、降低 glue code 脆弱度的團隊來說，這是很實用的一章。它讓你開始思考「流程」而不只是「單次訓練」，也更接近真實團隊的交付方式。\u003C\u002Fp>\u003Cul>\u003Cli>目標：自動化多步驟工作流\u003C\u002Fli>\u003Cli>結果：可重用的 pipeline 結構\u003C\u002Fli>\u003Cli>練習：完成 orchestration 作業\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>4. 部署不是只有一種做法\u003C\u002Fh2>\u003Cp>部署模組會比較 online、streaming 和 batch 三種方式，並提到 \u003Ca href=\"https:\u002F\u002Fflask.palletsprojects.com\u002F\" target=\"_blank\" rel=\"noopener noreferrer\">Flask\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Faws.amazon.com\u002Fkinesis\u002F\" target=\"_blank\" rel=\"noopener noreferrer\">AWS Kinesis\u003C\u002Fa>、Lambda 與 batch scoring。這種安排很重要，因為不同模型對延遲、吞吐和成本的要求完全不同。\u003C\u002Fp>\u003Cp>如果你主要只接觸過訓練，這一章會幫你補上「模型怎麼真的到使用者手上」的那一段。你會更清楚什麼情況該做 web service，什麼情況適合串流，什麼情況 batch 就夠了。\u003C\u002Fp>\u003Cul>\u003Cli>Online：web service 或 streaming\u003C\u002Fli>\u003Cli>Offline：batch scoring\u003C\u002Fli>\u003Cli>工具：Flask、AWS Kinesis、Lambda\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>5. 監控是 production ML 的分水嶺\u003C\u002Fh2>\u003Cp>這門課把 monitoring 當成正式模組處理，而不是附帶補充。它同時涵蓋 service monitoring 與 batch monitoring，工具包含 \u003Ca href=\"https:\u002F\u002Fprometheus.io\u002F\" target=\"_blank\" rel=\"noopener noreferrer\">Prometheus\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fwww.evidentlyai.com\u002F\" target=\"_blank\" rel=\"noopener noreferrer\">Evidently\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fgrafana.com\u002F\" target=\"_blank\" rel=\"noopener noreferrer\">Grafana\u003C\u002Fa>，以及 Prefect、MongoDB、Evidently 的組合。\u003C\u002Fp>\u003Cp>這種拆法很實際，因為線上服務和批次任務出問題的方式不一樣。若你想觀察 data drift、服務健康度、任務結果，這一章可以當作入門工具箱。\u003C\u002Fp>\u003Ccode>Web monitoring: Prometheus + Evidently + Grafana\nBatch monitoring: Prefect + MongoDB + Evidently\u003C\u002Fcode>\u003Ch2>6. 把工程習慣補齊，系統才撐得久\u003C\u002Fh2>\u003Cp>最後一個核心模組把焦點拉回工程實務，包括 unit test、integration test、linting、formatting、pre-commit hooks、\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ffeatures\u002Factions\" target=\"_blank\" rel=\"noopener noreferrer\">GitHub Actions\u003C\u002Fa> 的 CI\u002FCD，以及 \u003Ca href=\"https:\u002F\u002Fwww.terraform.io\u002F\" target=\"_blank\" rel=\"noopener noreferrer\">Terraform\u003C\u002Fa> 的 infrastructure as code。\u003C\u002Fp>\u003Cp>如果你已經有模型，卻總覺得交付品質不穩，這一段很值得補。它不是在教你多一個模型技巧，而是在教你怎麼讓流程可審查、可自動化、可重現，這正是 production 團隊最在意的部分。\u003C\u002Fp>\u003Cul>\u003Cli>測試：unit 與 integration\u003C\u002Fli>\u003Cli>自動化：pre-commit、GitHub Actions\u003C\u002Fli>\u003Cli>基礎設施：Terraform\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>怎麼挑，才不會白學\u003C\u002Fh2>\u003Cp>如果你想要的是一條免費、結構清楚、從實驗到部署再到監控的路線，這門課很適合。它特別適合已經會 Python 和 Docker、但還沒把 ML 送進 production 的人。\u003C\u002Fp>\u003Cp>如果你只想快速補單點技能，像是 MLflow、部署或監控，也可以直接跳到對應模組；但若你想最後拿出一個能展示的系統，跟著 final project 走完整條線會更划算。\u003C\u002Fp>","9 個免費模組、14.8k 星標，從實驗追蹤到部署監控與最終專案，幫你判斷這門 MLOps 課程是否適合把模型推進 production。","github.com","https:\u002F\u002Fgithub.com\u002FDataTalksClub\u002Fmlops-zoomcamp",null,"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781542984202-6g6y.png","industry","zh","8c10e73a-b4e7-444b-9a70-421823b16755",[17,18,19,20,21,22,23],"MLOps","MLflow","deployment","monitoring","CI\u002FCD","Terraform","production ML",[25,26,27],"免費 9 週課程，適合把模型訓練接到 production 流程","重點模組涵蓋實驗追蹤、pipeline、部署、監控與工程化","最適合已有 Python 與 Docker 基礎、想做端到端 MLOps 的讀者",0,"2026-06-15T17:02:28.556043+00:00","2026-06-15T17:02:28.55+00:00","da242733-a19a-4cb7-b706-05f8699aa19e",{"tags":33,"relatedLang":43,"relatedPosts":47},[34,36,38,39,41],{"name":35,"slug":19},"Deployment",{"name":17,"slug":37},"mlops",{"name":20,"slug":20},{"name":21,"slug":40},"cicd",{"name":18,"slug":42},"mlflow",{"id":15,"slug":44,"title":45,"language":46},"mlops-zoomcamp-path-to-production-ml-en","MLOps Zoomcamp maps the path to production ML","en",[48,54,60,66,72,78],{"id":49,"slug":50,"title":51,"cover_image":52,"image_url":52,"created_at":53,"category":13},"867b8247-e1b4-42cd-acb5-62caeeeea152","kalshi-adds-solana-perpetual-futures-after-xrp-zh","Kalshi 上架 Solana 永續合約","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781553773666-el0h.png","2026-06-15T20:02:30.33552+00:00",{"id":55,"slug":56,"title":57,"cover_image":58,"image_url":58,"created_at":59,"category":13},"63358330-a783-4029-a837-53fa4b33fd47","mlops-is-not-optional-for-production-ml-zh","想把 ML 用到生產環境，MLOps 不是選配","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781543880750-cdza.png","2026-06-15T17:17:22.084947+00:00",{"id":61,"slug":62,"title":63,"cover_image":64,"image_url":64,"created_at":65,"category":13},"fb1d2caa-dc25-4298-bde9-c53b0ff4502b","cloudflare-too-expensive-after-share-price-surge-zh","Cloudflare 漲太多了，現在買只是在接估值風險","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781539367968-dmjg.png","2026-06-15T16:02:18.514984+00:00",{"id":67,"slug":68,"title":69,"cover_image":70,"image_url":70,"created_at":71,"category":13},"7f4c85a1-7f7d-428c-875b-144bea2b8b34","turbovec-cuts-10m-vector-ram-to-4gb-zh","TurboVec 把 10M 向量壓到 4GB","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781528569742-vbog.png","2026-06-15T13:02:22.818062+00:00",{"id":73,"slug":74,"title":75,"cover_image":76,"image_url":76,"created_at":77,"category":13},"0d168fc7-0d4b-4653-aba4-1f058a075b7d","midjourney-v8-1-default-model-update-zh","Midjourney V8.1 變成預設模型，速度與細節都升級","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781515078543-4z93.png","2026-06-15T09:17:18.754939+00:00",{"id":79,"slug":80,"title":81,"cover_image":82,"image_url":82,"created_at":83,"category":13},"856138b5-19e2-4328-9637-ca9baa17e48f","midjourney-vs-zh","Midjourney 免費方案 vs 付費方案","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1781514187435-4dch.png","2026-06-15T09:02:34.997559+00:00",[85,90,95,100,105,110,115,120,125,130],{"id":86,"slug":87,"title":88,"created_at":89},"ee073da7-28b3-4752-a319-5a501459fb87","ai-in-2026-what-actually-matters-now-zh","2026 AI 真正重要的事","2026-03-26T07:09:12.008134+00:00",{"id":91,"slug":92,"title":93,"created_at":94},"83bd1795-8548-44c9-9a7e-de50a0923f71","trump-ai-framework-power-speech-state-preemption-zh","川普 AI 框架瞄準電力、言論與州權","2026-03-26T07:12:18.695466+00:00",{"id":96,"slug":97,"title":98,"created_at":99},"ea6be18b-c903-4e54-97b7-5f7447a612e0","nvidia-gtc-2026-big-ai-announcements-zh","NVIDIA GTC 2026 重點拆解","2026-03-26T07:14:26.62638+00:00",{"id":101,"slug":102,"title":103,"created_at":104},"4bcec76f-4c36-4daa-909f-54cd702f7c93","claude-users-spreading-out-and-getting-better-zh","Claude 用戶更分散，也更會用","2026-03-26T07:22:52.325888+00:00",{"id":106,"slug":107,"title":108,"created_at":109},"bd903b15-2473-4178-9789-b7557816e535","openclaw-raises-hard-question-for-ai-models-zh","OpenClaw 逼問 AI 模型價值","2026-03-26T07:24:54.707486+00:00",{"id":111,"slug":112,"title":113,"created_at":114},"eeac6b9e-ad9d-4831-8eec-8bba3f9bca6a","gap-google-gemini-checkout-fashion-search-zh","Gap 把結帳搬進 Gemini","2026-03-26T07:28:23.937768+00:00",{"id":116,"slug":117,"title":118,"created_at":119},"0740e53f-605d-4d57-8601-c10beb126f3c","google-pushes-gemini-transition-to-march-2026-zh","Google 把 Gemini 轉換延到 2026 年 3…","2026-03-26T07:30:12.825269+00:00",{"id":121,"slug":122,"title":123,"created_at":124},"e660d801-2421-4529-8fa9-86b82b066990","metas-llama-4-benchmark-scandal-gets-worse-zh","Meta Llama 4 分數風波又擴大","2026-03-26T07:34:21.156421+00:00",{"id":126,"slug":127,"title":128,"created_at":129},"183f9e7c-e143-40bb-a6d5-67ba84a3a8bc","accenture-mistral-ai-sovereign-enterprise-deal-zh","Accenture 攜手 Mistral AI 賣主權 AI","2026-03-26T07:38:14.818906+00:00",{"id":131,"slug":132,"title":133,"created_at":134},"191d9b1b-768a-478c-978c-dd7431a38149","mistral-ai-faces-its-hardest-year-yet-zh","Mistral AI 迎來最硬的一年","2026-03-26T07:40:23.716374+00:00"]