dist-tf.pdf$ whoami 藤原秀平 (FUJIWARA Shuhei) Twitter: @shuhei_fujiwara GitHub: @sfujiwara ▶ Google Developer Expert (Machine Learning) ▶ TensorFlow User Group Tokyo Organizer 2
Better performance with the tf.data API Â |Â TensorFlow Core
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR I0000 00:00:1723689002.526086 112933 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:17
MLIR: A new intermediate representation and compiler framework
Posted by The TensorFlow MLIR Team The TensorFlow ecosystem contains a number of compilers and optimizers that operate at multiple levels of the software and hardware stack. As a day-to-day user of TensorFlow, this multi-level stack might manifest itself as hard-to-understand compiler and runtime errors when using different kinds of hardware (GPUs, TPUs, mobile). These components, starting from th
Deep learning experiments in OCaml
Last year we held a machine learning seminar in our London office, which was an opportunity to reproduce some classical deep learning results with a nice twist: we used OCaml as a programming language rather than Python. This allowed us to train models defined in a functional way in OCaml on a GPU using TensorFlow. Specifically we looked at a computer vision application, Neural Style Transfer, and
Apache Hadoop上ã§TensorFlowã‚’å‹•ã‹ã™ã€ŒTonY〠| OSDN Magazine
ç±³Microsoft傘下ã®LinkedIn開発者ã¯9月12æ—¥ã€Apache Hadoop上ã§TensorFlowã‚’ãƒã‚¤ãƒ†ã‚£ãƒ–ã«å‹•ã‹ã™ã‚ªãƒ¼ãƒ—ンソースプãƒã‚¸ã‚§ã‚¯ãƒˆã€ŒTensorFlow on YARN（TonY）ã€ã‚’発表ã—ãŸã€‚ TensorFlow on YARN（TonY）ã¯ã€å¤§è¦æ¨¡ãªApache Hadoop実装上ã§åˆ†æ•£åž‹ã§æ©Ÿæ¢°å¦ç¿’ã‚’é‹ç”¨ã™ã‚‹ãŸã‚ã«LinkedIn社内ã§é–‹ç™ºã•ã‚ŒãŸãƒ•ãƒ¬ãƒ¼ãƒ ワーク。å˜ä¸€ãƒŽãƒ¼ãƒ‰ã¾ãŸã¯åˆ†æ•£åž‹ã®TensorFlowトレーニングをHadoopアプリケーションã¨ã—ã¦å‹•ã‹ã™ã“ã¨ãŒã§ãる。 開発ãƒãƒ¼ãƒ ã«ã‚ˆã‚‹ã¨ã€ãã‚Œã¾ã§ã€ŒTensorFlow on Sparkã€ã‚„Intelã®ã€ŒTensorFlowOnYARNã€ã‚’試ã—ãŸãŒã€ä¿¡é ¼æ€§ã‚„柔軟性ã«æ¬ ã‘ãŸãŸã‚æ–°ãŸã«é–‹ç™ºã™ã‚‹ã“ã¨ã«ã—ãŸã¨ã„ã†ã€‚TonYã§ã¯ã€ãƒªã‚½ãƒ¼ã‚¹ãƒã‚´ã‚·ã‚¨ãƒ¼ã‚·ãƒ§ãƒ³ã‚„コンテナ環境è¨å®šãªã©ã®ã‚¿ã‚¹ã‚¯å‡¦ç†ã‚’通ã˜ã¦TensorFlo
GitHub - nfmcclure/tensorflow_cookbook: Code for Tensorflow Machine Learning Cookbook
This chapter intends to introduce the main objects and concepts in TensorFlow. We also introduce how to access the data for the rest of the book and provide additional resources for learning about TensorFlow. General Outline of TF Algorithms Here we introduce TensorFlow and the general outline of how most TensorFlow algorithms work. Creating and Using Tensors How to create and initialize tensors i
Distributed TensorFlow ã§ãƒ‡ãƒ¼ã‚¿ä¸¦åˆ—を試ã—ã¦ã¿ã‚‹ - Qiita
データ並列ã¨ã¯ å‰å›žã¯Distributed TensorFlowã®ãƒ“ルドã‹ã‚‰ãƒ¢ãƒ‡ãƒ«ä¸¦åˆ—ã¾ã§ã‚’è¡Œã„ã¾ã—ãŸãŒã€ä»Šå›žã¯ãƒ‡ãƒ¼ã‚¿ä¸¦åˆ—ã«ã‚ˆã‚‹å¦ç¿’を試ã—ã¦ã¿ã¾ã™ã€‚並列化ã«ã¯ãƒ¢ãƒ‡ãƒ«ä¸¦åˆ—ã¨ãƒ‡ãƒ¼ã‚¿ä¸¦åˆ—ã®ï¼’種類ãŒã‚ã‚Šã¾ã™ãŒã€å¤§é›‘把ã«è¨€ã†ã¨ä¸‹è¨˜ã®ã‚ˆã†ã«ãªã‚Šã¾ã™ã€‚ モデル並列: データ1000個ã«å¯¾ã™ã‚‹å·¨å¤§ãªæ¼”ç®—1回を100人ã§åˆ†æ‹…ã™ã‚‹ データ並列: 1人ã‚ãŸã‚Šãƒ‡ãƒ¼ã‚¿10個ãšã¤å°åˆ†ã‘ã«ã—ã¦100人ã§åˆ†æ‹…ã™ã‚‹ モデル並列ã¯å½“然ãªãŒã‚‰ãƒ¢ãƒ‡ãƒ«ã«ä¾å˜ã™ã‚‹ã®ã§ã€ãƒ‡ãƒ¼ã‚¿ä¸¦åˆ—ã§ä¸€åº¦ã«æ‰±ã†ãƒ‡ãƒ¼ã‚¿ã‚’減らã™ã»ã†ãŒæ±Žç”¨æ€§ã¯é«˜ã„ã¨ã„ãˆã‚‹ã§ã—ょã†ã€‚ パラメータã®å…±æœ‰ å¦ç¿’ã«ãŠã‘るデータ並列化ã§ã¯ã€åŒã˜ãƒ‘ラメータをæŒã£ãŸãƒ¢ãƒ‡ãƒ«ã®ã‚³ãƒ”ーを複数作りã€ãƒãƒƒãƒã‚’å°åˆ†ã‘ã«ã—ã¦ãã‚Œãžã‚Œã®ãƒ¢ãƒ‡ãƒ«ã®ã‚³ãƒ”ーã«æ¸¡ã—ã€å„々ã«å¾®åˆ†ã‚’計算ã•ã›ã¾ã™ã€‚ã¤ã¾ã‚ŠåŒã˜ãƒ‘ラメータをもã£ãŸãƒ¢ãƒ‡ãƒ«ã‚’デãƒã‚¤ã‚¹ã”ã¨ã«æŒãŸã›ãªã‘ã‚Œã°ãªã‚‰ãªã„ã®ã§ã™ãŒã€ãã®ã‚ãŸã‚Šã®æ‰±ã„ãŒå°‘ã—ã‚ã‹ã‚Šã«ãã„ã§ã™
コピペã‹ã‚‰è„±å´ã—ã¦è‡ªç”±ã« Deep Learning ã™ã‚‹ãŸã‚ã® TensorFlow å†å…¥é–€ - Qiita
ã¯ã˜ã‚ã« TensorFlow ã¯ãªã‹ãªã‹ã¨ã£ã¤ãã«ãã„部分ãŒã‚ã‚‹ã‹ã¨æ€ã„ã¾ã™ã€‚書ãæ–¹ãŒç‹¬ç‰¹ãªã®ã§ã€æ…£ã‚Œã‚‹ã¾ã§ã«æ™‚é–“ãŒã‹ã‹ã‚‹ã‹ã¨æ€ã„ã¾ã™ã€‚å…¬å¼ã® MNIST ã¯ä¸€é€šã‚Šã‚„ã£ãŸã‘ã©ã€è‡ªåˆ†ã§è€ƒãˆãŸãƒ‡ã‚£ãƒ¼ãƒ—ニューラルãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ (DNN) ã‚’ã©ã†æ›¸ã„ãŸã‚‰ã„ã„ã‹ã‚ã‹ã‚‰ãªã„……ãªã‚“ã¦ã“ã¨ã‚‚ã‚ã‚‹ã®ã§ã¯ãªã„ã§ã—ょã†ã‹ã€‚ ã“ã®è¨˜äº‹ã§ã¯ã€ã‚·ãƒ³ãƒ—ルãªå•é¡Œã‚’é¡Œæã«ã—ã¦ã€ TensorFlow をコピペã«é ¼ã‚‰ãšã€è‡ªåˆ†ã®æ›¸ããŸã„ DNN ã‚’ 低レベル API を使ã£ã¦ã©ã®ã‚ˆã†ã«æ›¸ãã‹ã‚’最å°æ§‹æˆã§èª¬æ˜Žã—ã¦ã„ãã¾ã™ï¼ˆä»Šã•ã‚‰æ„Ÿã¯ã‚ã‚Šã¾ã™ãŒã€‚。）。ベースã¯å…¬å¼ãƒ‰ã‚ュメント㮠Low Level APIs ã®ã‚ãŸã‚Š ã§ã™ã€‚ 内容ã¨ã—ã¦ã¯ï¼š TensorFlow ã®æœ€å°æ§‹æˆè¦ç´ TensorFlow ã§ç·šå½¢é–¢æ•°ã®ãƒ•ã‚£ãƒƒãƒ†ã‚£ãƒ³ã‚° TensorFlow 㧠DNN を実装 ã®ã‚ˆã†ã«ãªã£ã¦ã„ã¾ã™ã€‚ã¾ãŸã€ä»¥ä¸‹ã®ã‚ˆã†ãªæ–¹ã‚’対象ã¨ã—ã¦ãŠã‚Š
p2 インスタンスã¸ã® TensorFlow å°Žå…¥ - Qiita
ã¯ã˜ã‚ã« ã“ã®è¨˜äº‹ã¯ AWS ã® GPU マシンã®ãªã‹ã§æœ€ã‚‚性能ã®é«˜ã„ p2 ç³»ã®ãƒžã‚·ãƒ³ã§ã®ã€ TensorFlow ã®ç’°å¢ƒæ§‹ç¯‰ã®ãƒ¡ãƒ¢ã§ã™ã€‚ 最終的㫠AWS Step Function 㨠AWS Lambda を組ã¿åˆã‚ã›ã¦ã€ã“ã“ã§ä½œã£ãŸãƒžã‚·ãƒ³ã‚’スãƒãƒƒãƒˆã‚¤ãƒ³ã‚¹ã‚¿ãƒ³ã‚¹ã¨ã—ã¦ç«‹ã¡ä¸Šã’自動ã§å¦ç¿’を回ã™ã“ã¨ã‚’目指ã—ã¦ã„ã¾ã™ã€‚ ※ p2 インスタンスã¯ç¾åœ¨æ±äº¬ãƒªãƒ¼ã‚¸ãƒ§ãƒ³ã«ã¯å°Žå…¥ã•ã‚Œã¦ã„ã¾ã›ã‚“。オレゴン・ãƒãƒ¼ã‚¸ãƒ‹ã‚¢ãªã©ã®ãƒªãƒ¼ã‚¸ãƒ§ãƒ³ã‚’使ã£ã¦ãã ã•ã„。 イメージã®é¸å®š TensorFlow å…¬å¼ãƒ‰ã‚ュメント ã‚’èªã‚€ã¨ Linux ã®å ´åˆ Ubuntu をベースã«æ›¸ã‹ã‚Œã¦ã„るよã†ã§ã™ã®ã§ã€ Ubuntu ã®ã‚¤ãƒ³ã‚¹ã‚¿ãƒ³ã‚¹ã‚’é¸ã³ã¾ã—ãŸã€‚ Ubuntu Server 16.04 LTS (HVM), SSD Volume Type 予゠CUDA 環境ãªã©ãŒå…¥ã£ã¦ã„ã‚‹ Amazon Linux ã® AMI ç‰
TensorFlowãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«MNIST(åˆå¿ƒè€…å‘ã‘)ã‚’TensorBoardã§è¦‹ãˆã‚‹åŒ– - Qiita
TensorFlowã®åˆå¿ƒè€…å‘ã‘ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«ã‚’TensorBoardã§å‡ºåŠ› TensorFlowã®åˆå¿ƒè€…å‘ã‘ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«MNIST For ML Beginnersã‚’TensorBoardã«å‡ºåŠ›ã—ã¦è¦‹ãˆã‚‹åŒ–ã—ã¾ã—ãŸã€‚ã§ã™ãŒã€TensorBoardã®è¦‹æ–¹(特ã«Graph以外)ã‚’ã»ã¨ã‚“ã©ç†è§£ã—ã¦ã„ãªã„ã®ã§ã‚³ãƒ¼ãƒ‰ã¨ç°¡å˜ãªè§£èª¬ã®ã¿ã§ã™ã€‚Deep Learningを実践ã—ã¦ã„ãã¨ã€éžå¸¸ã«æœ‰ç”¨ãã†ã§ã™ãŒã€ãã“ã¾ã§ã®ãƒ¬ãƒ™ãƒ«ã«é”ã—ã¦ã„ãªã„ã®ã§ãƒ»ãƒ»ãƒ» 環境:python3.5 tensorflow1.1 å‚考リンク TensorFlowã‚’Windowsã«ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ« Pythonåˆå¿ƒè€…ã§ã‚‚ç°¡å˜ã ã£ãŸä»¶ ã€å…¥é–€è€…å‘ã‘解説】TensorFlow基本構文ã¨ã‚³ãƒ³ã‚»ãƒ—ト ã€å…¥é–€è€…å‘ã‘解説】TensorFlowãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«MNIST(åˆå¿ƒè€…å‘ã‘) TensorFlow APIメモ ã€TensorBoard入門】T
åˆã‚ã¦ã®TensorFlow - イントãƒãƒ€ã‚¯ã‚·ãƒ§ãƒ³ã¨ã—ã¦ã®ç·šå½¢å›žå¸° - Qiita
（追記: 改訂版ã®è¨˜äº‹ã‚’投稿ã—ã¾ã—㟠- http://qiita.com/TomokIshii/items/0a7041ad337f68f71286 ） 先週(2015/11/9)，Deep Learninã®Framework ã®"TensorFlow"ãŒå…¬é–‹ã•ã‚ŒãŸãŒï¼Œãƒ‰ã‚ュメントã®èª¬æ˜Žã€ŒMNIST（手書ãæ•°å—ã®åˆ†é¡žï¼‰ã¯æ©Ÿæ¢°å¦ç¿’ã®"Hello World" ã§ã‚る．ã€ã¨ã„ã†ç®‡æ‰€ã«ç´å¾—ãŒã„ã‹ãªã„．Courseraã®Machine Learning (Stanford)ã§ã‚‚ãã†ã ã£ãŸãŒï¼Œæ©Ÿæ¢°å¦ç¿’ã‚’åˆæ©ã‹ã‚‰å¦ã¶å ´åˆï¼Œã‚„ã¯ã‚Šæœ€åˆã¯ Linear Regression（線形回帰）ã¨ï¼Œå€‹äººçš„ã«è€ƒãˆã‚‹ï¼Ž 本記事ã§ã¯ï¼Œæœ€åˆã«Linear Regression（線形回帰）ã®ã‚³ãƒ¼ãƒ‰ã‚’調ã¹ï¼Œæ¬¡ã« Logistic Regression（ãƒã‚¸ã‚¹ãƒ†ã‚£ãƒƒã‚¯å›žå¸°ï¼‰ã®ã‚³ãƒ¼ãƒ‰ã‚’作æˆã—ã¦ï¼ŒTensorFlowã®é›°å›²æ°—ã‚’ã¤ã‹
Dots deep learning部_20161221
dot DeepLearning部ã§ã®ç™ºè¡¨ã«ä½¿ã†è³‡æ–™ã§ã™ã€‚ ã“ã‚Œã¾ã§ã®ç„¼ãç›´ã—ã§ã™ãŒã€ã„ã‚ã„ã‚分ã‹ã‚Šã‚„ã™ããªã£ã¦ã„ã¾ã™ã€‚ TensorFlow ã®æ¦‚è¦ãŒåˆ†ã‹ã‚‹è³‡æ–™ãªã©ãŒè¿½åŠ ã•ã‚Œã¦ãŸã‚Šã—ã¾ã™ã€‚
Google Cloud Dataflowã§TensorFlowã®Grid Searchã‚’ã—ã¦ã¿ãŸ - Qiita
å°è¦æ¨¡ãªãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã§å¦ç¿’ã‚’ã•ã›ã‚‹æ™‚ã€å¤§ã¾ã‹ãªå‚¾å‘を見るãŸã‚ã«ãƒ¢ãƒ‡ãƒ«ã®ãƒ‘ラメータを振りãŸã„時ãŒã‚ã‚Šã¾ã™ã€‚scikit-learnã«ã¯Grid Searchã¨Cross Validationã‚’åŒæ™‚ã«è¡Œã†GridSearchCVã¨ã„ã†åå‰ãã®ã¾ã‚“ã¾ãªä¾¿åˆ©æ©Ÿèƒ½ãŒã‚ã‚Šã¾ã™ãŒã€ ã“れをTensorFlowã§ã‚‚〠リモートã§ä¸¦åˆ—ã«ã€ ãã—ã¦é¢å€’ãªåˆ†æ•£å‡¦ç†ã‚³ãƒ¼ãƒ‰ã‚’書ã‹ãšã« Jupyter Notebookã‹ã‚‰ç°¡å˜ã« ã‚„ã‚ŠãŸã„ã§ã™ã‚ˆã。 ã¯ã„ã€Google Cloud Dataflowãªã‚‰ãã‚ŒãŒã§ãã‚‹ã‚“ã§ã™ï¼ 準備 ã“ã“ã¯å‰å›žã®è¨˜äº‹ã¨ã»ã¼åŒã˜ã§ã™ãŒã€ãƒ¢ãƒ‡ãƒ«ã«ã‚ˆã£ã¦ã¯ãƒžã‚·ãƒ³ã‚¿ã‚¤ãƒ—を変ãˆãŸæ–¹ãŒè‰¯ã„ã§ã—ょã†ã€‚worker_options.machine_typeã®éƒ¨åˆ†ã§æŒ‡å®šã§ãã¾ã™ã€‚ ã¾ãŸnum_workersを指定ã™ã‚‹ã¨autoscaleãŒç„¡åŠ¹ã«ã•ã‚Œã‚‹ã‚ˆã†ã§ã™ã€‚今回ã¯6ã‚’è¨å®šã—ãŸã®ã§ã€workerãŒ6ã¤ä¸€æ°—ã«
ClassCat® AI Research – クラスã‚ャット – Agno, AgentOS, MCP, LangChain/LangGraph, CrewAI
ã“ã®ä¾‹ã¯ã€è¿½è·¡ã¨æ•´é “ã®å‘上ã®ãŸã‚ã«åå‰ä»˜ã Step オブジェクトを使用ã™ã‚‹ãƒ¯ãƒ¼ã‚¯ãƒ•ãƒãƒ¼ã‚’実演ã—ã¾ã™ã€‚ã“ã®ãƒ‘ターンã¯ã€å˜ç´”ãªã‚·ãƒ¼ã‚±ãƒ³ã‚·ãƒ£ãƒ«å®Ÿè¡Œã‚’ç¶æŒã—ãªãŒã‚‰ã€æ˜Žç¢ºãªã‚¹ãƒ†ãƒƒãƒ—è˜åˆ¥ã¨æ‹¡å¼µãƒã‚®ãƒ³ã‚°ã‚’æä¾›ã—ã¾ã™ã€‚ Agno 2….
GitHub - yahoo/TensorFlowOnSpark: TensorFlowOnSpark brings TensorFlow programs to Apache Spark clusters.
TensorFlowOnSpark brings scalable deep learning to Apache Hadoop and Apache Spark clusters. By combining salient features from the TensorFlow deep learning framework with Apache Spark and Apache Hadoop, TensorFlowOnSpark enables distributed deep learning on a cluster of GPU and CPU servers. It enables both distributed TensorFlow training and inferencing on Spark clusters, with a goal to minimize t
Windowsã§GPU使ã£ã¦TensorFlowã‚’å‹•ã‹ã™ãƒ¡ãƒ¢(TensorBoardã‚‚å‹•ã‹ã™) - Qiita
ã“ã“何年ã‹æ•°ãƒ¶æœˆãŠãã«ã€Œãªã‚“ã‹æµè¡Œã£ã¦ã‚‹ã—機械å¦ç¿’やらãã°ã€ã¨æ€ã£ã¦æ‰‹ã‚’ã¤ã‘ã‚‹ã‘ã©ã™ãã«é “挫ã™ã‚‹ã¨ã„ã†ã®ã‚’ç¹°ã‚Šè¿”ã™ã†ã¡ï¼Œç¶šã‹ãªã„åŽŸå› ã®ã²ã¨ã¤ãŒå®Ÿè¡Œæ™‚é–“ã ã¨æ°—ã¥ãã¾ã—ãŸï¼Žè‡ªåˆ†ä»¥å¤–ã«ã‚‚実際ã«å‹•ãã‚‚ã®ã‚’触りãªãŒã‚‰ã˜ã‚ƒãªã„ã¨å‹‰å¼·ã®ãƒ¢ãƒãƒ™ãƒ¼ã‚·ãƒ§ãƒ³ã‚’ä¿ã¦ãªã„人ã¯å®Ÿè¡Œæ™‚é–“ã§è‹¦æˆ¦ã—ã¦ã„ã‚‹ã‹ã¨æ€ã„ã¾ã™ï¼Ž GPU使ã†ã¨10å€ãらã„高速化ã™ã‚‹ã‚‰ã—ã„ã®ã§ä½¿ã„ãŸã„ãªã¨æ€ã£ã¦ã„ãŸã¨ã“ã‚ã§ï¼ŒTensorFlowãŒWindowsã«å¯¾å¿œã—ã¦ã„ãŸã®ã§ï¼Œã²ã¨ã¾ãšæ™®æ®µä½¿ã£ã¦ã„ã‚‹Windowsノートã§å®Ÿè¡Œã—ã¦ã¿ã¾ã—ãŸï¼Ž TensorFlow 0.12 ㌠Windows をサãƒãƒ¼ãƒˆ CPU/GPU/AWSã§ã®Tensorflow実行速度比較 準備 環境． Windows 10(64bit) Python 3.5 (今回ã¯æœ¬å®¶ã®ã‚„ã¤ã‚’使ã„ã¾ã—ãŸï¼ŽAnacondaç‰ã®ãŠå¥½ããªç‰©ã‚’) NVIDIAã®GPU CUDA + cuDNN 使ãˆ
分散å¦ç¿’用TensorFlowコードã®æ›¸ãæ–¹ - ã‚ã‚‚ã‚ã‚‚
何ã®è©±ã‹ã¨ã„ã†ã¨ Google Cloud MLを利用ã—ã¦ã€TensorFlowã®åˆ†æ•£å¦ç¿’ã‚’è¡Œã†æ–¹æ³•ã§ã™ã€‚å–り急ãŽã€è‡ªåˆ†ç”¨ã®ãƒ¡ãƒ¢ã¨ã—ã¦å…¬é–‹ã—ã¦ãŠãã¾ã™ã€‚ 分散å¦ç¿’ã«ã¯ã„ãã¤ã‹ã®ãƒ‘ターンãŒã‚ã‚Šã¾ã™ãŒã€æœ€ã‚‚シンプルãªã€Œãƒ‡ãƒ¼ã‚¿åˆ†æ•£ã€ã®å ´åˆã‚’説明ã—ã¾ã™ã€‚å„ノードã¯åŒã˜ãƒ¢ãƒ‡ãƒ«ã«å¯¾ã—ã¦ã€å€‹åˆ¥ã«å¦ç¿’データをé©ç”¨ã—ã¦ã€Variableã‚’ä¿®æ£ã™ã‚‹å‹¾é…ベクトルを計算ã—ã¾ã™ã€‚ãã‚Œãžã‚Œã§è¨ˆç®—ã—ãŸå‹¾é…ベクトルを用ã„ã¦ã€å…±é€šã®Variableã‚’ä¿®æ£ã—ã¦ã„ãã¾ã™ã€‚ å‰æçŸ¥è˜ TensorFlowã®åˆ†æ•£å¦ç¿’処ç†ã‚’è¡Œã†éš›ã¯ã€3種類ã®ãƒŽãƒ¼ãƒ‰ã‚’使用ã—ã¾ã™ã€‚ ・Parameter Server：WorkerãŒè¨ˆç®—ã—ãŸå‹¾é…ベクトルを用ã„ã¦ã€Variableã®ã‚¢ãƒƒãƒ—デートを行ã„ã¾ã™ã€‚ ・Worker：教師データã‹ã‚‰å‹¾é…ベクトルを計算ã—ã¾ã™ã€‚ ・Master：Workerã¨åŒæ§˜ã®å‡¦ç†ã«åŠ ãˆã¦ã€å¦ç¿’済ã¿ãƒ¢ãƒ‡ãƒ«ã®ä¿å˜ã‚„テストセットã«å¯¾ã™ã‚‹
ランã‚ング
ランã‚ング
メンテナンス
リリースã€éšœå®³æƒ…å ±ãªã©ã®ã‚µãƒ¼ãƒ“スã®ãŠçŸ¥ã‚‰ã›
最新ã®äººæ°—エントリーã®é…ä¿¡
j次ã®ãƒ–ックマーク
kå‰ã®ãƒ–ックマーク
lã‚ã¨ã§èªã‚€
eコメント一覧を開ã
oページを開ã
Yahoo supercharges TensorFlow with Apache Spark | TechCrunch
TFUG#3 Rettyã«ãŠã‘ã‚‹Deep Learningã®è‡ªç„¶è¨€èªžå‡¦ç†ã¸ã®å¿œç”¨äº‹ä¾‹ - Speaker Deck
GitHub - kevinhughes27/TensorKart: self-driving MarioKart with TensorFlow
{{#tags}}- {{label}}
{{/tags}}