ä»ã®LLMãåãå·»ãç¶æ³ã«ã¤ãã¦ç´¹ä»ãã¾ãã
ã¯ããã«ä»¥ä¸ã®ãªã¼ãã³ãªããã¸ã§ã¯ãã®ä¸ç°ã§ï½¤å¤§è¦æ¨¡è¨èªã¢ãã«ããã«ã¹ã¯ã©ããã§ä½ãç·´ç¿ããã¾ã。24å¹´3æç¾å¨ï½¤ååè ãåéä¸ã§ã。 ãªãã¸ããªå½è©²ããã¸ã§ã¯ãã®æ¨æºã³ã¼ããå ¬éâ»ãããã®ã§ï½¤ãããèµ°ããã¦ã¿ã¾ã。 â»24/3/5æç¹ã§ï½¤ã¾ã ã¬ãã¸ããªå ã«ï½¤å·¥äºä¸ã®ç®æãå¤ã 、ããã¾ã。 ãã®ãªãä¸ã§ã¯ï½¤äºåå¦ç¿ã¼äºå¾å¦ç¿ã¼è©ä¾¡ã¾ã§ï½¤ä¸æ°é貫(?)ãããã¤ãã©ã¤ã³ãæä¾ããã¦ãã¾ãâ»ï½¡ 0. ç°å¢æ§ç¯ããã¸ã§ã¯ãã®æ¬çªç°å¢ã¯ã¯ã©ã¦ãã§ãã、ä»åã¯ææã¡ã®ubuntuã使ãã¾ã。 Dockerã¯ãæ軽ãªä¸æ¹ã§ï½¤ã¹ãã³ã³ä¸ã§ä½¿ãã¨ï½¤ã©ãããé度ãè½ã¡ããããã¨ã®åãèããã®ã§ï½¤condaã§ä½ãã¾ã(ã¨ã¯ãã、pipãã使ããªãã®ã§ï½¤pyenvã§ããããã¨æãã¾ã)。 å¿ è¦ãªãã·ã³é©å½ãªlinux: ä¾ãã°Ubuntu 22.04.3 LTS GPU: 20 GBãããã¯æ¬²ãããã? ãã£ã¹ã¯
次ã ã¨çºè¡¨ããããªã¼ãã³ãªæ¥æ¬èªå¤§è¦æ¨¡ã¢ãã« ã©ããªã£ã¦ããã®ï¼ï¼ã¨ããæãã§ãããã æããnpakaããã¯ããã£ããGoogle Colabã§åããã¦ããã£ãããã¾ãã ãã ãGoogle Colabã ã¨æ¯åã¢ãã«ã®ãã¦ã³ãã¼ãã大å¤ãªã®ã§ããã¼ã«ã«ã§Docker使ã£ã¦æ軽ã«åãããã¨ãããªãã¨ãããã¨ã§ãã£ã¦ã¿ã¾ããã 以ä¸GitHubã®ãªãã¸ããªã«Dockerfileã¨ãµã³ãã«ããã°ã©ã ãããã¦ãã¾ãããã£ããã£ã½ããã¨ãã§ããããã«ãã¦ãã¾ãã ä¸è¨ã§ããµã¤ãã¼ã¨ã¼ã¸ã§ã³ãã¨ãªã³ãã®LLMã両æ¹åãã¾ãã 使ç¨ç°å¢ åæã¨ãªãç°å¢ã§ãã使ç¨ãã¦ããPCã®ã¹ããã¯ã¯ä»¥ä¸ã§ãã é ç® å 容
GWä¸ã®é£ç¶ä¼ç»ã¨ãã¦ãç»åçæAIãMidjourneyãã使ã£ãAIã¤ã©ã¹ãå¶ä½TIPSè¨äºããå±ããã¦ãã¾ããã¾ãã5æ2æ¥ã«ã¯ãããã«é¢é£ããçæ¾éä¼ç»ãå®æ½ãã¾ããããã覧ããã ããã§ããããï¼ ãã¡ãã®æ¾éã§ã¯ãæ¬é£è¼ä¼ç»ã®ç¬¬1åã¨ç¬¬2åã§åãä¸ããããMidjourneyãã®å§ãæ¹ã®ç´¹ä»ã¨ããMidjourneyãã使ã£ãä¼¼é¡çµµå¶ä½ã«å®éã«ãã£ã¬ã³ã¸ãã¦ããã®ã§ãããæ¾éæéã®å¶ç´ããããæéå ã«ã§ãããã£ãã¢ã¹ãã¼ç·¨éé¨ã¤ã°ãã®ä¼¼é¡çµµãã¡ãã£ã¨å¾®å¦ãªæãã§ããã ããã§ä»åã¯ãå¼ãç¶ãããã®ã¨ãä¾é¡ã¨ãã¦å©ç¨ããåçã使ã£ã¦ãããã¡ãã£ã¨æ¬äººã«ä¼¼ãã¾ã§è©¦è¡é¯èª¤ãç¹°ãè¿ããããã©ããªçµæã«ãªãã®ãããããã®è©¦è¡é¯èª¤ã®éç¨ã¨å ±ã«ãå±ããããã¨æãã¾ããçæ¾éãæªè¦è´ã®æ¹ã¯ãåç»ã®å¾åé¨åã48å20ç§ããããã®æ åãè¦ã¦ãããæ¬ç¨¿ããèªã¿ããã ããã¨ããã«æ¥½ãããã¨æãã¾ãã®ã§
æ¨å¹´æ«ããæ¥æ¿ã«è©±é¡ãå¼ãã ChatGPTããã®å é¨çãªãã¼ã¸ã§ã³ã¢ããã§ããGPT-3.5ã¨GPT-4ã¯ããããMicrosoftããå¼·åãªæ¯æ´ãåããç±³OpenAIã¨ããä¼æ¥ãç¬å çã«æä¾ãã大è¦æ¨¡è¨èªã¢ãã«ï¼LLMï¼ã ã ããã«å¼å¿ãããã®ããã«ãMeta社ããã¯LLaMAï¼ã©ãï¼ããªãªã¼ã¹ãããLLaMAãChatGPTã¨ã®1ä¸3ååã®ä¼è©±ãã¼ã¿ã§å¾®èª¿æ´ããAlpacaï¼ã¢ã«ãã«ï¼ãAlpacaãShareGPTï¼GPTã¨ã®ä¼è©±ãæå¿ããªã¼ãã³ã«ãããã®ï¼ã®ãã¼ã¿ã§å¾®èª¿æ´ããVicunaï¼ãã¯ã¼ãã£ï¼ã¨ãã£ãæ´¾çã¢ãã«ã次ã ã¨ç»å ´ããã ããããLLaMAã¯Meta社ã®æåã«ãããã¢ã«ãããã¯ç¨ééå®ãã¨ããç¸ãããããã¾ãGPTã®åºåã«ã¯OpenAIã®å©ç¨è¦å®ã§ãGPTã®åºåçµæãå ã«GPTã«å¯¾æã§ããå¼·åãªAIãä½ã£ã¦ã¯ãããªããã¨ããå¶ç´ããããããAlpacaãVicuna
ãå°å ¥ã ã»ãã®è¨äºã®æçµç®æ¨ã¯ããªã¢ã«ã¿ã¤ã å¤æã§ããã ãããã¤ã¹ã«ãªã£ã¦Discordã§é話ãããã§ãã ã»å¤§ã¾ããªè¨äºã®æµãã¯ãRVCã§å¦ç¿ã¢ãã«ãä½ããâãå¦ç¿ã¢ãã«ãVC Clientã«å ¥ãããâããªã¢ã«ã¿ã¤ã ãã¤ãã§ã³å®æï¼ãã§ãã ã»ãRVCãã¨ããè¶ åãéãªã¢ã«ã¿ã¤ã é³å£°å¤æã½ããã話é¡ã«ãªã£ãã¨æã£ãããç¿æ¥ã«ãVC Clientãã¨ããè¶ åãã½ããããªã¢ã«ã¿ã¤ã é³å£°å¤æã«å¯¾å¿ããã®ãèããã®ã§ãæ¥ãã§è¨äºãæ¸ãã¾ããã ã»ãã®è¨äºã§ã¯ããã ãããã¤ã¹ã«ãªããã¨ãç®æ¨ã¨ãã¦ãã¾ãããå¦ç¿å ã®é³å£°ããããã°ã©ããªé³å£°ã«ããªãã¾ãã ãæé ã â ï½RVCã®ãã¦ã³ãã¼ãï½ ã»ä»¥ä¸ã®URLãéãã¦ããRVC-beta.7zããDLãã¾ããï¼è¦Hugging Faceã¢ã«ã¦ã³ãï¼ï¼Hugging Faceã¢ã«ã¦ã³ããæã£ã¦ãªã人ã¯ç¡æãªã®ã§ä½ãã¾ãããï¼ https://huggi
ããã«ã¡ã¯ãTuringæ ªå¼ä¼ç¤¾ã®æ©æ¢°å¦ç¿ãã¼ã ã§ã¤ã³ã¿ã¼ã³ããã¦ããä¹å·å¤§å¦ä¿®å£«1å¹´ã®å²©æ¿(@colum2131)ã§ãã Turingã¯å®å ¨èªåé転EVã®éçºãããã¹ã¿ã¼ãã¢ããã§ãã èªåé転æè¡ã«ããã¦ãã«ã¡ã©ãã»ã³ãµæ å ±ã¯æ£ç¢ºãªå¶å¾¡ãããä¸ã§ä¸å¯æ¬ ãªè¦ç´ ã§ããä¸æ¹ãèªç¶è¨èªããã«ãã¢ã¼ãã«ãªå¦çãå¿ è¦ã¨ãªãç¶æ³ããã°ãã°åå¨ãã¾ããç¹ã«å®å ¨èªåé転è»ã«ããã¦ã¯ãé³å£°èªèã«ãã£ã¦ãã©ã¤ãã¼ã®é³å£°å½ä»¤ãèªèãããªã¢ã«ã¿ã¤ã ã§é©åãªå¶å¾¡ãè¡ããã¨ããè¤éãªäº¤éç¶æ³ã®èæ¯ã«ããã³ã³ãã¯ã¹ããç解ãããã¨ãã£ãè¦æ±ãåºã¦ãã¾ãããã®ããã«ããåºç¤ã¢ãã«ãã¨å¼ã°ãããããªãèªç¶è¨èªå¦çãå«ã大è¦æ¨¡ã¢ãã«ã®å¦ç¿ãå¿ è¦ã«ãªã£ã¦ããã¨ãç§ãã¡ã¯èãã¦ãã¾ãã ããã§æ¬è¨äºã§ã¯ãè¨èªã¢ãã«ã®çºå±ã®æµãã¨Turingãç®æãåºç¤ã¢ãã«ã®éçºã«ã¤ãã¦ç´¹ä»ãã¾ãï¼ 1. è¨èªã¢ãã«ã¨ã¯ï¼ è¿å¹´ã®è¨èªå¦çã¢ã
ãGoogle Colabãã§ãFlexGenãã試ããã®ã§ã¾ã¨ãã¾ããã ã注æããOPT-30Bãã®ãã£ãããã¢ãå®è¡ããã«ã¯ããGoogle Colab Pro/Pro+ãã®ããã¬ãã¢ãã®ããã¤ã¡ã¢ãªãã¨ã300GBã»ã©ã®Googleãã©ã¤ããå¿ è¦ã§ãã 1. FlexGenãFlexGenãã¯ãéãããGPUã¡ã¢ãª (16GB T4 ã¾ã㯠24GB RTX3090ãªã©) ã§å¤§è¦æ¨¡è¨èªã¢ãã«ãå®è¡ããããã®é«ã¹ã«ã¼ãããçæã¨ã³ã¸ã³ã§ãã 大è¦æ¨¡è¨èªã¢ãã« (LLM) ã¯ãChatGPTãCopilotãªã©ã®ã¢ããªã±ã¼ã·ã§ã³ã®ã³ã¢ã«ãªãã¾ãããLLMæ¨è«ã®è¨ç®ã¨ã¡ã¢ãªã®è¦ä»¶ãé«ããããå¾æ¥ã§ã¯è¤æ°ã®ãã¤ã¨ã³ãã¢ã¯ã»ã©ã¬ã¼ã¿ã使ç¨ããå ´åã«ã®ã¿å®è¡å¯è½ã§ããããFlexGenãã¯ãLLMæ¨è«ã®ãªã½ã¼ã¹è¦ä»¶ã1ã¤ã®ã³ã¢ãã£ãã£GPU (T4ã3090ãªã©) ã«ã¾ã§ä¸ãããã¾ãã¾ãªãã¼ã
人工ç¥è½ãå¦ã¶ããã®ãã¼ãããã ãã®ãã¼ã¸ã§ã¯ã人工ç¥è½ã深層å¦ç¿ãå¦ãã ãã¨ã®ãªãæ¹ã対象ã«ã ããããå¦ã¶ããã®ãã¼ãããããç´¹ä»ãã¦ãã¾ãã æ¬ãã¼ããããã§ã¯éæç®æ¨ã¨ãã¦ã ãç 究è ãããã¼ã¿ãµã¤ã¨ã³ãã£ã¹ãããã¨ã³ã¸ãã¢ãããã¸ãã¹ãã® 4ã¤ã®è·æ¥ãã¨ã«4ã¤ã®ã¬ãã«ãè¨ãã¦ãã¾ãã ã¾ãã¯ã¬ãã«0ã¨ãã¦ã人工ç¥è½ã«ã¤ãã¦ã®åºç¤çãªç¥èãå¦ã³ã¾ãããã æè¡ã«è§¦ããï¼å¦ç¿æ³å®æéï¼1æéï¼ ã¾ãã¯ææ°ã®AIæè¡ã«è§¦ãã¦ï¼AIã«ãã£ã¦ã©ã®ãããªãã¨ãã§ããã®ãã確èªãã¦ã¿ã¾ãããï¼ ã¡ã¸ã£ã¼ãªãµã¼ãã¹ãå«ãããã¤ãç´¹ä»ãã¾ãï¼ ã»ChatGPT ChatGPTâ¨ãã§ã«å©ç¨ãããã¨ãããæ¹ãå¤ããã¨æãã¾ãããOpenAIãéçºã»éå¶ãã大è¦æ¨¡è¨èªã¢ãã«ï¼LLMï¼ãã£ãããããã§ããLLMã§ã¯ä»ã«Googleã®GeminiãAnthropicã®ClaudeãMistral AIãªã©
TL;DR Chrome Extensionçµç±ã§ç¬èªã«å¦ç¿ããMLã¢ãã«ã社å é å¸ã§ããããã«ãã¾ãã ã¢ãã«ã¯ã¦ã¼ã¶ã®ãã©ã¦ã¶ä¸ã§å®è¡ããã®ã§ä½è¨ãªéä¿¡ãçºçããã¯ã©ã¦ã代ãä¸è¦ã§ã èæ¯ ããã«ã¡ã¯ãCADDi AI Lab MLOpsãã¼ã ã®ä¸æéµä»ã§ããæ®æ®µã¯æ©æ¢°å¦ç¿ã¨ã³ã¸ãã¢ãã¼ã ã®ä½ãã¢ãã«ãVertex Endpointsã使ç¨ãã¦APIã¨ãã¦æä¾ãããããã¤ãã©ã¤ã³ã«çµã¿è¾¼ãã§æ¨è«çµæããã¼ã¿æä¾ããããããä»äºããã¦ãã¾ããã¢ãã«ã¯æ§ã ãªç¨®é¡ãããã¾ãããä¸çªå¤ãã®ã¯å³é¢ç»åããç¹å®ã®å¤ãæ¨è«ããããä½ããã®ã¯ã©ã¹ã«åé¡ãããããªã¢ãã«ã§ãã ãã®ãããªä¸ã§ãAPIæä¾ããã¨ãµã¼ã代ãããããã¦ã¼ã¶ã«API使ã£ã¦ãããã®ãã¡ãã£ã¨æéãããããªããã¨èãããã¨ãããããµã¨ãChrome extensionã§MLã¢ãã«ãæä¾ãã¡ããã°ãã¦ã¼ã¶ã¯extensionãå ¥ããã
骨è£éãæ³å¾äºåæå¼è·å£«ãå é£åºç¥ç財ç£æ¦ç¥æ¨é²äºåå±åäºå®è£ä½ãã¨ã³ã¿ãã¤ã³ã¡ã³ãæ³åãå°éã主ã«ã¢ãã¡ãã²ã¼ã ãAIããããããVRæ¥çã®æ¹ããµãã¼ããã¦ãããå®ã¯ãã©ãã¼ã§ããããï¼Twitterï¼@hajime_ideiï¼ AIã使ãå´ãæ°ãã¤ããªããã°ãããªãã㨠ã½ãªï¼ 便å©ãªAIãã¼ã«ãç¶ã ã¨ç»å ´ããä¸ãAIãä½åã¥ããã«åãå ¥ããæ¹ãå¢ããã¨æããã§ãããã ãAIåµä½ç©ã«ã¯èä½æ¨©ãçºçããªããªã©ãé常ã«ãããããåé¡ãããã¨ãããã¨ã§ããã ã§ã¯ãå®éã«æã ã¯ãªã¨ã¤ã¿ã¼ãAIãã¼ã«ãæ´»ç¨ãã¦ããå ´åããã©ãã«ãé¿ããããã«ã¯ã©ããªãã¨ã«æ°ãã¤ããã°ããã®ã§ããããã åºäºï¼ ã¾ããèä½ç©æ§ã®åé¡ã«ã¤ãã¦ã§ãããèªåã®èä½ç©ã¨ãã¦ä¿è·ãããã®ã§ããã°ãAIãçæããä½åã«ã¯å ¨ä½çã«äººã®æãå ãããã¨ãéè¦ã ã¨èãã¦ãã¾ãã ãããããã¨ã§ããAIãçæããä½åã ããèä½æ¨©ã¯çº
ããã«ã¡ã¯ã社å ãã¼ã¿ãµã¤ã¨ã³ã¹ãã¼ã YAMALEXã®@Ssk1029Takashiã§ãã æè¿ã¯RTX4090ã®ãã·ã³ãè²·ã£ã¦é»æ°ä»£ãä¸ãã£ã¦ãã¦ãã¾ãã æ¨ä»NLPçéã§ã¯äºåå¦ç¿ã¢ãã«ãåºã¦ããã¯ãæ¤ç´¢ã¨ããã®ããã¼ã¯ã¼ãã§ã®æ¤ç´¢ã ãã§ã¯ãªããæç« ãå ¥åã«ãã¦ä¼¼ããããªæç« ãæ¢ãé¡ä¼¼ææ¸æ¤ç´¢ã使ãããããã«ãªãã¾ããã ãããªä¸ã§ãä»åã¯FAQã対象ã«ããé¡ä¼¼ææ¸æ¤ç´¢ãSentence BERTã使ã£ã¦è©¦ãã¦ã¿ã¾ãã FAQã§ããããå°ãã㨠ä»åããã㨠Sentence BERTã¨ã¯ æ¤è¨¼ FAQãã¼ã¿ã»ããããæç« ãã¯ãã«ãå¦ç¿ãã åçæããæç« ãã¯ãã«ãçæãã 質åæããè¿ããåçãè¦ã¤ãã åèªãæºãã¦ãããæãã«åçãåå¾ã§ããã®ãæ¤è¨¼ ã¾ã¨ã FAQã§ããããå°ãã㨠FAQã¯ããããå ´é¢ã§éè¦ãªæ å ±æºã§ããããã¾ãã¡æ¤ç´¢ããã¥ããæ å ±ã§ãããã¾ãã FAQãå©ç¨ã
12æ10æ¥ã®2022ã½ããã¦ã§ã¢ãã¹ãã¢ããã³ãã«ã¬ã³ãã¼ã§ãã Launchable社ã§ã¨ã³ã¸ãã¢ã¨ãã¦åãã¦ããcvuskã¨ç³ãã¾ããæ©æ¢°å¦ç¿çéã§ã¯æ©æ¢°å¦ç¿ãå®ç¨åããããã®ã·ã¹ãã éçºã®æ¬ãæ¸ãã¦ããã¾ããããè¯ãã£ããèªãã§ã¿ã¦ãã ããã ãæ©æ¢°å¦ç¿ã·ã¹ãã ãã¶ã¤ã³ãã¿ã¼ã³ã ãæ©æ¢°å¦ç¿ã·ã¹ãã æ§ç¯å®è·µã¬ã¤ãã æ¬ããã°ã§ã¯æ©æ¢°å¦ç¿ãç¨ãã¦ãã¹ãå®è¡ãå¹çåããææ³ã¨ãã¦ãPredictive Test Selectionã«ã¤ãã¦èª¬æãã¾ãããã¹ãå®è¡æéãã³ã¹ãã§èª²é¡ãæ±ãã¦ããã¨ã³ã¸ãã¢ã«å½¹ã«ç«ã¤ã¨å¹¸ãã§ãã æ¨ä»ã®éçºã«ããããã¹ãäºæ 2002å¹´ã«ããã¹ãé§åéçºããä¸ã«åºã¦ãã½ããã¦ã§ã¢éçºã§ãã¹ããæ¸ããã¨ã常èã«ãªã£ã¦æ©20å¹´ãçµã£ã¦ãã¾ãããã®éã«ã¯ã©ã¦ãã®ç»å ´ãDevOpsã®æ®åã«ããããã¹ããCI/CDãã¤ãã©ã¤ã³ã§èªåå®è¡ããã³ã¼ãã¨ãããã¯ãå質ãç¶æã
ã©ããããã¹ã¦ã®çµæ¸æ´»åãããã¸ã¿ã«åãããç¦å³¶ã§ãã æ¬æ¥ã¯ãLayerXãè³ãããSaaS+Fintechãã¨ããæ°ããæ½®æµã«ã¤ãã¦ã®è§£èª¬ãæã ã®èããç´¹ä»ã§ããã°ã¨æãã¾ãã ãã®è¨äºã§ãããããã«ãSaaS+Fintechãã¨ç¹ã«ç¸æ§ã®è¯ãé åã§ããæ¯åºç®¡çã®DXãé¢é£ãã¦ãã話ã§ãã SaaS+Fintechã¯ç¬¬ï¼ä¸ä»£ã®ã½ããã¦ã§ã¢ãã¸ãã¹ã¢ãã«SaaS+Fintechã¨ããæ°ããæ½®æµãSaaS+Fintechãã¨ã¯ç±³å½ã®èåVCã§ããa16zã2020å¹´8æã«æ稿ããFintech Scales Vertical SaaSã¨ããè¨äºã«ã¦æã¡åºãããæ¦å¿µã§ãããã以æ¥ã½ããã¦ã§ã¢ãã¸ãã¹ã®æå 端ã®æ½®æµã¨ãã¦èªèããã¦ãã¾ãã https://future.com/fintech-scales-vertical-saas/ ããã½ããã¦ã§ã¢ãã¸ãã¹ã¢ãã«ã®é²åã®æ´å²ã¯ã 第1ä¸ä»£(
ããã«ã¡ã¯ã å¨å® ã®æ©ä¼ãå¢ãã¦ä»¥æ¥Youtubeãè¦ãæ©ä¼ãå¢ããæ©æ¢°å¦ç¿ãªã©ãåå¼·ã§ãããã£ã³ãã«ãããã¤ãæ¢ãã¦ã¯è¦ã¦ãã¾ãããæ¢ããä¸ã§ããã£ãã¨æã£ããã®ãã¡ã¢ãã¦ããã®ã§ããããã£ãããªã®ã§å ¬éãããã¨æãã¾ããæ¥æ¬èªã®ã½ã¼ã¹ããããã®ã®ã¿å¯¾è±¡ã«ãã¦ããããããç¡æã§ããã®ãï¼ãã¨æã£ããã£ã³ãã«ãç´¹ä»ãããã¨æãã¾ãã主観ã§ä»¥ä¸ã®ã¬ãã«ã«åãã¾ãããããã¾ã§åèç¨åº¦ã«ãé¡ããããã¾ãã åºæ¬ï¼Pythonã触ã£ã¦ã¿ã人 Pythonã®èª¬æã»åããæ¹ãªã©ã解説ãã¦ãã¦ãåç»ã«ãã£ã¦ã¯è¸ã¿è¾¼ãã å 容ã«ãªã å¿ç¨ï¼ã¢ã«ã´ãªãºã ã使ãããªããã人 ãmodel.fit(X, y)ãã¦åããã¦ã¿ãããããè¸ã¿ãã¿ãPythonèªä½ã®èª¬æã¯å°ãªã çºå±ï¼ç 究éçºãããã人 ææ°ã®ææ³ã®ä»çµã¿ã®ç解ãªã©ã主ç¼ã§ãããPythonã®è§£èª¬ã¯ã»ã¼ç¡ã ããããããã®ãã£ã³ãã«ãããã¾ããããã²ã³
Google ãå ¬éãã¦ãããããè¯ããã¼ã¿åæã®ããã®ã¬ã¤ãããã¯ãGood Data Analysisãã§ããã¼ã¿åæã®è¦æãç°¡æ½ã«ã¾ã¨ãããã¦ãã¦æåãã 2022-03-08 Google ã®éå ¬å¼ããã°ã§ãThe Unofficial Google Data Science Blog ã¨ãããã¼ã¿ãµã¤ã¨ã³ã¹ããã¼ãã«ããããã°ãããã ãã®ä¸ã§ã Practical advice for analysis of large, complex data sets ã®è¨äºãå ã«ãã¦ä½ããã Google Developers Guides: Machine Learning Guides > Good Data Analysis ãæ¨æ¥è¦ããã¦èªãã§ãããç´ æ´ãããããã¥ã¡ã³ãã ã£ãã®ã§ãããã§ãã®æåãå ±æãããã£ãã®ã§çãã¨ã£ããã ãã Good Data Analysis ã®æ¦
ããã«ã¡ã¯ãTURINGæ ªå¼ä¼ç¤¾ã§ã¨ã³ã¸ãã¢ããã¦ãã ããã¤ã ã§ãã ãã®è¨äºãèªãã§ããã ãã¦ããçæ§ã®å¤ãã¯TURINGã£ã¦ï¼ãããªæãã ã¨æãã¾ãã ååãç¥ã£ã¦ããã ãã¦ããæ¹ã§ããæè¿Twitterã¨ãã§èããã©çµå±ãªã«ãã£ã¦ãã®ï¼å®éãªã«ããã£ãã®ï¼ã¨ããæãããããã¾ããã TURINGã¯ã2021å¹´8æã«åµæ¥ããå®å ¨èªåé転EVã®è£½é ã»éçºãç®æãä¼ç¤¾ã§ãã ãã®è¨äºã§ã¯ãTURINGæåã®ããã¸ã§ã¯ããBudapestã ã«ã¤ãã¦ã話ããã¾ãã æ§æã¯ã ï¼. æ¦è¦ ï¼. æè¡çã¢ããã¼ã ï¼. ãã®å¾ã®åãçµã¿ ï¼. çµããã« ä»é²ï¼Budapestã®ååã®ç±æ¥ ã§ãã 1. æ¦è¦ Budapestã¯ãã«ã¡ã©ç»åãã深層å¦ç¿ã§End-to-Endã§ãã³ãã«ãå¶å¾¡ããéå®ç©ºéã®æ±ºã¾ã£ãã«ã¼ããå¨åãç¶ããã¨ããããã¸ã§ã¯ãã§ãã ãã®ããã¸ã§ã¯ãã®ç®çã¯ãè»ã®æ¹é ãã
ãªã»ãAIãä½ãå§ããæ¥ã®ã㨠ããã¯2021å¹´4æã®ãã¨ãä»æãè¿ãã°å¶ç¶ãéãªã£ã¦èµ·ããåºæ¥äºã§ããã 第ä¸ã®å¶ç¶ã¯ãã²ã¼ã AI(ã²ã¼ã ãèªåãã¬ã¤ããAI)ä¸ç4é£è¦ã®æ¹ã«ãªããã²ã¼ã AIã®åæ©çãªè©±ã30åç¨åº¦æãã£ã¦ãããã¨ã§ãã 第äºã®å¶ç¶ã¯ãTwitterã®ç¥ãåããããªã»ãã½ããRTAããªã競æããã£ã¦ããã®ãç®ã«ãããã¨ã§ãããªãããããããã¨æã£ãç§ã¯ããã«ããã®ç«¶æã ããªã»ãã§éã¶ãã©ãããã©ã¼ã ãã©ãã ãæ©ãä½ãããã競ããã®ã ã¨ãããã¾ããã é¢ç½ãããã£ã¦ã¿ããã YouTubeã§ã©ã¤ãé ä¿¡ããªãããç§ã¯ãªã»ãã½ããRTAããã£ã¦ã¿ã¾ããããã®æã®ã©ã¤ãã¯ãã¡ãã3æéã§å®æã§ããã°è¯ãã¨æã£ã¦ããã®ã§ãããæã£ããããæ©ãçµãã£ã¦ãã¾ãã¾ããã ãªã»ãAIã§ãä½ããã ãããã¦ç§ã®ãªã»ãAIå¶ä½ãéå§ãã¾ããã ä½ããããè¯ããããããªãã£ã ãªã»ãAIãä½
ã©ã³ãã³ã°
ã©ã³ãã³ã°
ã©ã³ãã³ã°
ãªãªã¼ã¹ãé害æ å ±ãªã©ã®ãµã¼ãã¹ã®ãç¥ãã
ææ°ã®äººæ°ã¨ã³ããªã¼ã®é ä¿¡
å¦çãå®è¡ä¸ã§ã
j次ã®ããã¯ãã¼ã¯
kåã®ããã¯ãã¼ã¯
lãã¨ã§èªã
eã³ã¡ã³ãä¸è¦§ãéã
oãã¼ã¸ãéã
{{#tags}}- {{label}}
{{/tags}}