Upgrade to Pro — share decks privately, control downloads, hide ads and more …

get-started-with-machine-learning-on-aws-20170401

 get-started-with-machine-learning-on-aws-20170401

IoT ALGYAN @ 20170401

1e5a15f4dc65c207a04a1e82a3f92e92?s=128

ryo nakamaru

April 01, 2017
Tweet

Transcript

  1. AWS Ͱ࢝ΊΑ͏ʂ͸͡Ίͯͷػցֶश IoT ALGYAN @ 2017.04.01

  2. @pottava (AWS Certified) SA, DevOps Engineer Pro ❤ Amazon ECS,

    AWS Batch, IAM
  3. גࣜձࣾεϐϯϑ

  4. ɹɹػցֶशͷ͓͞Β͍ • ػցֶशͱ͸ • ػցֶशͱਂ૚ֶश ɹɹֶश • σʔλͷ४උ • ࢼߦࡨޡ

    / POC • ֶश ɹɹਪ࿦ ɹɹAWS ར༻ Tips ࠓ೔͓࿩͢͠Δ͜ͱ 4
  5. ػցֶशͷ͓͞Β͍

  6. ػցֶश

  7. Կ͔ಛఆͷ໰୊͕͋Δͱͯ͠ɻ ίϯϐϡʔλʹ ٬؍తࣄ࣮ͷΈ Λ༩͑Δ͜ͱͰ ۩ମతͳճ౴ΛಘΔɺ·ͨ͸ͦΕΛվળ͢Δ͜ͱɻ ػցֶशͱ͸ 7

  8. y = ax + b a, b Λਓ͕ؒࣄલʹܾΊΔͷ͕Ұൠతϓϩάϥϛϯάɻ ࣮σʔλ͔Βίϯϐϡʔλʹܭࢉͤ͞Δͷ͕ػցֶशɻ ػցֶशͱ͸

    8
  9. y = ax + b ྫʣ x : ࠷ۙ஌Γ߹ͬͨਓͷಛ௃ y

    : ͜ͷͻͱͱকདྷ݁ࠗͨ͠Β޾ͤʹͳΕΔ͔ ػցֶशͱ͸ 9
  10. y = ax + b ྫʣ ਓؒʮ y = 0.7

    * ੑ֨x + 0.2 * ֎ݟx + 0.1 * ऩೖx ͰΑΖʯ ػցʮաڈͷσʔλ͔Β͍͑͹ ɹɹɹy = 0.4 * ੑ֨x + 0.1 * ֎ݟx + 0.5 * ऩೖx ͕ద੾ʯ ػցֶशͱ͸ 10
  11. y = ax + b a, b ΛܾΊΔͨΊͷ࡞ۀΛֶशɺ ܾ·ͬͨ a,

    b ͷ͜ͱΛֶशࡁΈϞσϧͱݴ͏ɻ ֶशࡁΈϞσϧΛ࢖ͬͯ ࣮ࡍʹ x Λ౤ೖ͠ y ΛಘΔͷ͕ਪ࿦ɻ ֶशͱਪ࿦ 11
  12. y = ax + b Ϗδωε্ॏཁͳͷ͸ɺ༏Εͨਪ࿦͕Ͱ͖Δ͔ɻ ༏Εͨਪ࿦Λ͢ΔͨΊʹ͸ɺ༏ΕͨϞσϧ͕ඞཁɻ ༏Εͨ a, b

    ΛܾΊΔͨΊͷֶश͕ɺ࿹ͷݟͤͲ͜Ζɻ ֶशͱਪ࿦ 12
  13. y = ax + b σʔλ͔Β a, b ΛٻΊΔํ๏͸ͨ͘͞Μ͋Δɻ ղ͖͍ͨ໰୊ʹΑͬͯɺద੾ͳํ๏ΛબͿඞཁ͕͋Δɻ

    ֶशΞϧΰϦζϜ 13
  14. y = ax + b σʔλ͔Β a, b ΛؼೲతʹٻΊΔ۩ମతͳํ๏ͷ͜ͱɻ •

    A / B Ͳͬͪʁ໰୊ → ϩδεςΟοΫճؼ • ച্Λ༧ଌ͍ͨ͠ → ઢܗճؼ • ސ٬Ληάϝϯτ෼͚͍ͨ͠ → k ฏۉ๏ • ϨίϝϯυΛग़͍ͨ͠ → ڠௐϑΟϧλϦϯά • … ֶशΞϧΰϦζϜ 14
  15. IoT ͰूΊͨσʔλ͔Β༏ΕͨϞσϧΛ࡞Γਪ࿦͢Δɻ ͖ͬ͞ͷֶशΞϧΰϦζϜɺ࢖͑ͦ͏Ͱ͢ΑͶʁ • ΋͏͙͢ނো͢Δʁ͠ͳ͍ʁ • ऩ֭ྔ͸Ͳͷ͘Β͍ʹͳΔͩΖ͏ʁ • ࣅ௨ͬͨάϧʔϓʹ෼͚͍ͨ •

    ௥Ճ஫จͯ͘͠Εͦ͏ͳαΠυϝχϡʔ͸ԿͩΖ͏ʁ IoT × ػցֶश 15
  16. Ұํɺࠓ೔΋࿩୊ͷ

  17. ਂ૚ֶश

  18. σΟʔϓϥʔχϯάɻ ଟ૚ߏ଄ͷωοτϫʔΫΛ༻͍ͨػցֶशͷ͜ͱɻ ྫʣ 4 ૚ͷωοτϫʔΫྫ ਂ૚ֶशͱ͸ 18

  19. ੨ؙʹ஋Λೖྗ͢Δͱɺ྘ؙʹ౴͕͑ग़ྗ͞ΕΔɻ ྫʣ͍҆ɺඒຯ͍͠ → ങ͏΂͖ 0.9ɺങΘͳ͍΂͖ 0.1 ਂ૚ֶशͱ͸ 19

  20. 2 ૚໨Ҏ߱ͷ֤ϊʔυ˓͕ɺલͷ૚͔ΒͷೖྗΛجʹ y = ax + b Λ࢖ͬͯࣗ෼ࣗ਎ͷ஋Λܭࢉɻ ਂ૚ֶशͱ͸ 20

  21. ΑΓ΋ͬͱ΋Β͍͠౴͑Λฦͨ͢Ίʹ ֤ϊʔυͷ a, b ΛࣄલʹܾΊΔͷ͕ɺֶशɻ શϊʔυͷ a, b ͕ܾ·Ε͹ɺͦΕֶ͕शࡁΈϞσϧɻ ਂ૚ֶशͱ͸

    21
  22. ػցֶशશൠɺղܾ͍ͨ͠໰୊͝ͱʹબ୒͢Δͷ͕ ֶशΞϧΰϦζϜɻਂ૚ֶश΋ྫ֎Ͱ͸ͳ͍ɾɾ ྫʣ • ͜ͷ͖Ύ͏Γ͸ S? M? L? → ৞ΈࠐΈχϡʔϥϧωοτϫʔΫ

    (CNN) • ࠓͷൃݴ͸ϙδςΟϒʁ → ࠶ؼܕχϡʔϥϧωοτϫʔΫ (RNN) • ਓ͕ඈͼग़͖ͯͨ͠ʂंΛݮ଎ʂʂ → ͋Ε͜Ε૊Έ߹Θͤ • … ਂ૚ֶशΞϧΰϦζϜ 22
  23. ඒ͍͠૊Έ߹ΘͤͷྫɺAlphaGoɻ ʢ࿦จ: http://www.nature.com/nature/journal/v529/n7587/abs/nature16961.html ʣ • 4 ͭͷσΟʔϓχϡʔϥϧωοτϫʔΫΛར༻ • ͦΕͧΕ ॳڃऀ

    pπ / தڃऀ pσ / ্ڃऀ pρ / ৘੎൑அ vθ ͱ͍ͬͨҐஔ෇͚ • pπ ͸ਓؒͷ 800 ສͷ൫໘σʔλΛݩʹֶशɻਫ਼౓͸௿͍͕ߴ଎ʹղΛಘΔɻ • pσ ͸ 13 ૚ͷ CNNɻ3,000 ສ൫໘Λ 50 GPU Ͱ 3.4 ԯεςοϓɺ3 िֶؒशɻ ϓϩͷࢦ͠खΛ 57.0% ͷਫ਼౓Ͱ༧૝Ͱ͖Δɻ • pρ ͸ 50 GPU Ͱ 1 ೔͔͚128 ສճࣗݾରઓɻطଘιϑτʹ 85% ͷѹ౗తউ཰ɻ • vθ ͸ pσ ͰϥϯμϜʹ 3,000 ສ൫໘Λੜ੒͠ɺpρ Ͱ 1 ԯ 6,000 ສճϩʔϧΞ΢τͨ͠উ཰Λ ڭࢣσʔλʹɺ50 GPU ͰҰिؒ 5,000 ສճ֬཰ޯ഑߱Լ๏Λ࣮ࢪɻ • ࣮ରઓͰ͸1,202 CPU + 176 GPU͕࢖ΘΕɺpσ Ͱ࣍ͷखબ୒ɺvθ Ͱ൫໘ධՁɻ • উ཰͸͍͍͕ཧ٧Ίͷ pρ ΑΓɺਓؒͷบΛֶΜͩ pσ Λ࢖͔ͬͨɻ • উҼ͸ pρ ͷపఈతͳڧԽֶशʹՃ͑ɺϞϯςΧϧϩ໦୳ࡧͱ CNN ͷ૊߹͔ͤɻ ਂ૚ֶशΞϧΰϦζϜ 23
  24. ࣮ફతͳωοτϫʔΫɺ ࣮ࡍʹ͸ͲΕ͘Β͍σΟʔϓͳϨΠϠʔͳͷʁ ྫʣ ৞ΈࠐΈχϡʔϥϧωοτϫʔΫͷҰछɺResNet ͸ 152 ૚ ਂ૚ֶशΞϧΰϦζϜ https://research.googleblog.com/2016/08/improving-inception-and-image.html 24

  25. ͓ɺ͓͏ɾɾ Ϟσϧͱܾͯ͠ΊΔม਺ͲΕ͚ͩ͋Δͷɾɾ 
 ֶशʹͲΕ͚͔͔ͩ࣌ؒΔͷɾɾ ͱ͍͏͔ɺϨΠϠʔఆٛ͢Δ͚ͩͰ৺͕ંΕͦ͏ɻ ࣗલͰ࣮૷ʁ·͋ແཧͰ͢ΑͶɾɾ ਂ૚ֶशΞϧΰϦζϜ 25

  26. ͦ͜Ͱ

  27. ਂ૚ֶशϑϨʔϜϫʔΫ

  28. ར༻ऀ͸ΞϧΰϦζϜͷ࣮૷Λ͢Δ͜ͱͳ͘ ֤छύϥϝλͷࢦఆ͚ͩͰֶशɾਪ࿦͕Ͱ͖Δɻ ྫʣTensorFlow ʹΑΔ৞ΈࠐΈχϡʔϥϧωοτϫʔΫ (CNN) ఆٛ ɹ ͲΜͳॱংͰͲΜͳ૚Λܦ༝͢Δ͔ɺ௚ײతʹΘ͔Δ ਂ૚ֶशϑϨʔϜϫʔΫ 28

  29. ΞϧΰϦζϜͷ࣮૷͸ͦͷಓͷϓϩʹ͓೚ͤͭͭ͠ɾɾ ࢲͨͪ͸΍Γ͍ͨ͜ͱ͚ͩͰ͖Δ࣌୅ɺ౸དྷɻ ྫʣը૾ʹ͍ࣸͬͯΔਓ͕஌Γ͍ͨ → TensorFlow Ͱ CNN Λ࢖͑͹ֶशɾਪ࿦Ͱ͖Δʂ ≒ Golang

    Ͱ HTTP/2 Λ࢖͑͹ηΩϡΞͳ௨৴΋؆୯ʂ ਂ૚ֶशϑϨʔϜϫʔΫ 29
  30. TensorFlowɺMXNetɺCaffeɺChainerɺTheanoɾɾ ͦΕͧΕͷಛ௃ΛؑΈͯͲΕΛ࢖͏ͷ͔ɻ • ରԠΞϧΰϦζϜ • ಈ࡞୺຤ɾ؀ڥ • ܭࢉ଎౓ / Ϧιʔεར༻ޮ཰

    • ར༻Մೳͳݴޠ / खଓతɾએݴత • εέʔϥϏϦςΟ / ෳ਺ GPUɺฒྻαʔόରԠ • ৘ใͷ๛෋͞ / ΤίγεςϜ / ঎༻αϙʔτ • … ਂ૚ֶशϑϨʔϜϫʔΫ 30
  31. ΫΠζͰ͢

  32. ʮ2 ͔݄લ͔ΒूΊͨσʔλ͔Β ໨ඪମॏ·Ͱ͋ͱԿϲ݄ ͔͔Δ͔༧ଌ͍ͨ͠ʯ

  33. ࣍ͷͲΕΛ࢖͏ɾɾʁ

  34. ϩδεςΟοΫճؼ ઢܗճؼ ৞ΈࠐΈχϡʔϥϧωοτϫʔΫ Ҩ఻తϓϩάϥϛϯά

  35. ࣮ફ͢Δલͷɺେ੾ͳϙΠϯτ

  36. ʮ͋ͳͨͷۀ຿ʹػցֶशΛ׆༻͢Δ 5 ͭͷϙΠϯτʯ https://www.slideshare.net/shoheihido/5-38372284 גࣜձࣾ Preferred Infrastructure ൺށ কฏ͞Μ ͱͯ΋͍͍εϥΠυͰͨ͠ɻ

    36
  37. ͯ͞

  38. ֶशͷྲྀΕΛ࠶֬ೝ

  39. ػցֶशͷྲྀΕ 2. σʔλલॲཧ 3. ֶश 4. ਪ࿦ 1. σʔλऩू 39

  40. Ҏ߱ɺ͜ͷྲྀΕʹԊ͍ɺAWS ΛͲ͏࢖͑͹͍͍ͷ͔ ར༻λΠϛϯάͱ໨తผʹ͝঺հ͠·͢ɻ ػցֶशͷྲྀΕ 2. σʔλલॲཧ 3. ֶश 4. ਪ࿦

    1. σʔλऩू 40
  41. ֶश

  42. σʔλͷ४උ

  43. 1ɹσʔλͷऩू ΋ͪΖΜɺIoT ͔ΒಘΒΕΔηϯασʔλ͸༗༻ʂʂ ͱ͸͍͑ɺ·ͣػցֶशΛࢼͯ͠ΈΔ͚ͩͳΒ Ұൠެ։͞ΕͨσʔλΛ׆༻͢Δͷ͕؆୯Ͱ͢ɻ 43

  44. Ұൠެ։͞Εͨը૾ू ݚڀ΋ίϯςετ΋੝ΜͰɺͨ͘͞Μ͋Γ·͢ɻ • MNIST ɹ http://yann.lecun.com/exdb/mnist/ • CIFAR-10 & CIFAR-100

    ɹ https://www.cs.toronto.edu/~kriz/cifar.html • ImageNet ɹ http://www.image-net.org/ • … 44
  45. AWS Public Datasets https://aws.amazon.com/jp/public-datasets/ 45

  46. Public Datasets | Google Cloud Platform https://cloud.google.com/public-datasets/ 46

  47. Public data sets for Azure analytics https://docs.microsoft.com/en-us/azure/sql-database/sql-database-public-data-sets 47

  48. Datasets « Deep Learning http://deeplearning.net/datasets/ 48

  49. http://www.data.go.jp/data/dataset 49 ͜Μͳͷ΋͋Δ

  50. AWS

  51. ؔ࿈αʔϏε܈

  52. σʔλऩूʹศརͳαʔϏε • AWS IoT • Amazon Kinesis Streams • Amazon

    CloudWatch Logs • Amazon S3 • Amazon DynamoDB • Amazon Cognito + AWS SDK • Amazon API Gateway 52
  53. 2ɹσʔλͷલॲཧ Python ͚ͩͰࡁΉͳΒͦΕͰ͍͍΋ͷͷɺ ෳ਺ͷσʔλιʔε͔ΒϝλσʔλΛऔಘͨ͠Γ େن໛ͳϑΝΠϧ͔ΒσʔλΛൈ͖ग़͢ͳΒ ઐ༻ͷιϑτ΢ΣΞ΍αʔϏε͕ศརɻ 53

  54. AWS ͷؔ࿈αʔϏε܈

  55. σʔλલॲཧʹศརͳαʔϏε • Amazon Mechanical Turk • Amazon Athena • AWS

    Lambda / Step Functions • AWS CloudWatch Events • Amazon EMR / Batch / EC2 55
  56. Amazon Mechanical Turk 56 ΞϝϦΧͰ͸ͦͷར༻͕ ͱͯ΋ྲྀߦ͍ͬͯΔ ͱͷ͜ͱɾɾ

  57. ࢼߦࡨޡ / POC

  58. ࢼߦࡨޡʹศརͳ΋ͷͨͪ

  59. ओʹՊֶٕज़ܭࢉ΍ػցֶशͷۀքͰ ͋Ε͜Εࢼߦࡨޡͨ͠ΓɺͦΕΛ୭͔ͱڞ༗͢ΔͨΊͷ 
 πʔϧɻଟ͘ͷݚڀऀ΍ΤϯδχΞʹѪ༻͞Ε͍ͯΔɻ git ͳͲͰόʔδϣϯ؅ཧ͢Δͷ΋༰қʂ Jupyter notebook 59

  60. ֶशʹ͸ͱͯ΋͕͔͔࣌ؒΔ΋ͷɻ ߦྻܭࢉ͕ಘҙͳ GPU Λ࢖͑͹͕࣌ؒઅ໿Ͱ͖·͢ʂ ࣗ୐ͷ PC ʹ͍ࢗͬͯ͞Δ GPU ͕࢖͑Δ͔΋ɾɾʁ (NVIDIA)

    GPU 60 ʢ͜Ε͕ࢗͬͯ͞Δਓ͸͍ͳ͍ͱࢥ͏͚Ͳ..ʣ
  61. (NVIDIA) GPU 61 ݱࡏ AWS Ͱ GPU Λ࢖͏ͱ͖ͷ Tips Λ·ͱΊ·ͨ͠

    https://speakerdeck.com/pottava/tesorflow-v1-dot-0-on-ec2
  62. ࢼߦࡨޡ͢Δʹ͸͜Ε΋ͱͯ΋ศརͰ͢ɻ ϥΠϒϥϦ͕ͲΜͲΜόʔδϣϯߋ৽ͯ͠΋େৎ෉ʂ Ϋϥ΢υ্ʹֶशɾਪ࿦Λ࣋ͬͯߦ͘ͱ͖ʹ΋༗༻ʂ docker run -it --rm -p 8888:8888 jupyter/tensorflow-notebook

    Docker 62
  63. Docker 63 ݱࡏ AWS Ͱ Docker Λ࢖͏ͱ͖ͷ·ͱΊ https://speakerdeck.com/pottava/containers-on-aws

  64. NVIDIA ͷ Docker Πϝʔδ 64 ҎԼͷΑ͏ͳܧঝؔ܎ͷΠϝʔδ͕ެ։͞Ε͍ͯ·͢ɻ ػցֶशͷಈ࡞ཁ݅ʹదͨ͠ΠϝʔδΛϕʔεʹɻ ಠࣗ Docker ΠϝʔδͷϏϧυ΋Ͱ͖·͢ɻ

    cuda:7.x-runtime ubuntu:14.04 cuda:7.x-devel cuda:7.x-cudax-runtime cuda:7.x-cudax-devel caffe (v0.14) digits (v4.0)
  65. AWS Ϣʔβ Retty ͞Μ͕ GPUʢ෺ཧʣΛങͬͨ࿩ɻ ͱͯ΋ڵຯਂ͍ɻ Ϋϥ΢υ͸ඞཁͳͷʁ 65 http://qiita.com/taru0216/items/dda1f9f11397f811e98a

  66. ࣮ࡍɺνʔϜͰݚڀɾ։ൃΛ͢ΔͱͳΔͱ ؀ڥͷ෼཭ / ෳ੡ / ڞ༗ / ݖݶ؅ཧͳͲ՝୊͸ͨ͘͞Μɻ ޙड़ͷ EMR

    or ECS + IAM ͳͲͷ૊Έ߹ΘͤΕ͹ ࠷৽ͷ GPU ؀ڥΛࣗಈ഑෍͢Δͱ͍ͬͨ͜ͱ΋ʂ AWS Ͱͷ෼ੳ؀ڥ 66
  67. AWS ͷؔ࿈αʔϏε܈

  68. ࢼߦࡨޡϑΣʔζʹศརͳαʔϏε • Amazon Machine Learning • Amazon EMR / EC2

    ‣ p2 / g2 (GPU) instances ‣ Deep Learning AMI • Amazon EBS / S3 / ECR 68
  69. ೋ߲෼ྨɺෳ਺Ϋϥε෼ྨɺઢܗճؼͷϚωʔδυαʔ ϏεɻσʔλͷऔΓࠐΈ͔Βɺֶशɾਪ࿦͕ߦ͑·͢ɻ αʔόͷ؅ཧ͕ෆཁͳͨΊɺεέʔϥϏϦςΟ΍ਪ࿦αʔ ϏεͷՄ༻ੑ͸ؾʹͤͣ OKʂ Amazon Machine Learning 69

  70. GPU Πϯελϯε 70 AWS ʹ͸ 2 छྨ͋Γ·͢ʢݱߦੈ୅ʣ g2 ܥ: NVIDIA

    GRID K520 ɹɹɹɹ1,536 CUDA cores / GPU ͕ 2 ͭͰ 1 ͭͷ K520 ɹɹɹɹg2 Ͱ࢖͑Δ GPU ͸ຊདྷάϥϑΟοΫɾήʔϛϯά༻్ p2 ܥ: NVIDIA Tesla K80 ɹɹɹɹഒਫ਼౓ԋࢉ࠷େ 2.91 TFLOPSɺ୯ਫ਼౓ԋࢉ࠷େ 8.74 TFLOPS ɹɹɹɹ2,496 CUDA cores / GPU ͕ 2 ͭͰ 1 ͭͷ K80 ɹɹɹɹp2 ͷ GPU ͸൚༻ίϯϐϡʔςΟϯά༻్
  71. Amazon DeepLearning AMI 71 શ෦ೖΓ AMIɺ͋Γ·͢ʂʂ TensorFlow 1.0, MXNet, Caffe,

    CNTK, Theano, Torchɻ CUDA 7.5, cuDNN 5.0, Anaconda ΋ɻ
  72. ֶश

  73. 3ɹֶश ΋ͬͱ΋ॏཁɺ͔ͭ΋ͷ͕͔͔࣌ؒ͘͢͝Δͱ͜Ζɻ ࣗࣾͷϞσϧΛ࡞Δͷ͸ͱͯ΋େม͕ͩ ΦϦδφϧͷ΋ͷ͕Ͱ͖Ε͹وॏͳ஌ࡒʹɻ Ϋϥ΢υͷ༷ʑͳαʔϏε͕αϙʔτͯ͘͠Ε·͢ɻ 73

  74. AWS ͷؔ࿈αʔϏε܈

  75. ֶशϑΣʔζʹศརͳαʔϏε • Amazon EMR / Batch / EC2 ‣ p2

    / g2 (GPU) instances ‣ Deep Learning AMI ‣ Spot Fleet / AutoScaling Group • Amazon Machine Learning • Amazon EFS / EBS / S3 / ECR • Amazon SQS 75
  76. AWS Batch 76 https://www.youtube.com/watch?v=UR8BI2Exkbc Պֶٕज़ܭࢉɾϋΠύϑΥʔϚϯείϯϐϡʔςΟϯά ༻్ͰਅՁΛൃش͢Δɺେن໛ͳεέʔϧɺδϣϒͷґ ଘఆ͕ٛՄೳͳϚωʔδυฒྻ෼ࢄόονॲཧج൫ɻ

  77. ͢Ͱʹ Black Belt ͷࢿྉ͕ެ։͞Ε͍ͯ·͢ɻ AWS Batch http://aws.typepad.com/sajp/2017/02/aws-black-belt-online-seminar-aws-batch.html 77

  78. ࢲ΋Ϣʔβࢹ఺Ͱݱঢ়Λ·ͱΊ·ͨ͠ɻ AWS Batch http://qiita.com/pottava/items/d9886b2e8835c5c0d30f 78

  79. ਪ࿦

  80. 4ɹਪ࿦ ֶशࡁΈͷϞσϧΛ࢖͍ɺਪ࿦͢Δɻ Ϗδωεͱ௚݁͢Δ͜ͱ͕ଟ͘ɺՔಇ͸ 24 / 365ɻ Մ༻ੑͱϨΠςϯγ͕ॏཁͳͷ͸ҰൠαʔϏεಉ༷ɻ ΋͔ͯ͠͠αʔόϨεͰ΋ɾɾ͍͚Δɾɾɾʁ 80

  81. AWS ͷؔ࿈αʔϏε܈

  82. ਪ࿦ϑΣʔζʹศརͳαʔϏε • Amazon ECS / EC2 ‣ p2 / g2

    (GPU) instances ‣ Deep Learning AMI ‣ Spot Fleet / AutoScaling Group • AWS Lambda / Amazon API Gateway • AWS ElasticBeanstalk • Amazon EFS / EBS / S3 / ECR 82
  83. ϑϧϚωʔδυͳ Docker ίϯςφΫϥελ؀ڥɻ GPU ϕʔεͷਪ࿦ΞϓϦέʔγϣϯͩͬͯಈ͖·͢ʂ Amazon ECS 83 https://speakerdeck.com/ayemos/build-image-classification-service-with-amazon-ecs-and-gpu-instances ΫοΫύουגࣜձࣾ

    છ୩ ༔Ұ࿠͞Μ
  84. ͪ͜Β΋͢Ͱʹ Black Belt ͷࢿྉ͕ެ։͞Ε͍ͯ·͢ɻ Amazon ECS 84 http://aws.typepad.com/sajp/2017/02/aws-black-belt-online-seminar-aws-batch.html

  85. AWS Lambda 85 αʔόϨεͰ MXNet ʹΑΔਪ࿦Λ͢Δ࣮૷ྫ http://aws.typepad.com/sajp/2017/01/ seamlessly-scale-predictions-with-aws-lambda- and-mxnet.html

  86. AWS ར༻ Tips

  87. ػցֶशΛ AWS Ͱ΍ΔͳΒɺ஌ͬͯಘ͢Δػೳ No. 1ʂ Ծ૝αʔόΛ҆͘࢖͑Δىಈํ๏ɻ AWS ͷσʔληϯλͷʮ༨৒෼ʯΛ ͜ͷֹۚͳΒ࢖͍·͢ʂͱʮೖࡳʯͯ͠ىಈɻ Spot

    Fleet / Spot Πϯελϯε 87
  88. ͜Ε·Ͱ঺հ͖ͯͨ͠ AWS ͷ֤αʔϏεΛ Ͳ͏࢖͍͍͔ͨΛ yaml / json Ͱએݴతʹهड़͓͖ͯ͠ Ұؾʹੜ੒ɾഁغ͢Δ͜ͱ͕Ͱ͖Δɻ ؆୯ਝ଎ʹɺ҆৺ͯ͠؀ڥ͕ߏஙͰ͖Δɻ

    ӡ༻ෛՙԼ͕Γ·͢ɻΠϯϑϥΛόʔδϣϯ؅ཧՄೳʹɻ CloudFormation 88
  89. Let’s try, anyway!

  90. ؼͬͨΒࣗ෼ͰҰ࿈ͷྲྀΕΛܦݧʂʂ࣮ફେࣄɻ ʮ਺ࣈը૾൑ఆ with TensorFlow on AWSʯ ࠓ೔ͷ॓୊ http://qiita.com/pottava/items/2fb2572f7099d432ebd9 90

  91. ػցֶशʹͲͷΫϥ΢υΛ࢖͏͔ʁͱߟ͑ΔΑΓ ৄ͍͠ਓΛั·͑ͯɺͬ͞ͱ৘ใΛूΊͯ ͍͍ͱ͜ͲΓͰ࢖͏ͷ͕Α͍ͱࢥ͍·͢ɻ 
 ͱ͍͏͔ɺࢼ͚ͩ͢ͳΒ ϩʔΧϧʹ؀ڥΛ੔͑Ε͹े෼Ͱ͢ɾɾ ·ͱΊ͡Όͳ͍ɺ·ͱΊ 91

  92. JAWS-UG AI ࢧ෦

  93. ίϯςϯπ • AWS Ͱ AI αʔϏεΛ࣮૷ɾӡ༻͢ΔͨΊͷ ɹҰൠతͳٕज़৘ใɺ஌ݟɺࣄྫڞ༗ͷ৔ • ͢Ͱʹ׆༻͍ͯ͠Δํ •

    ಋೖΛݕ౼͍ͯ͠Δํ • ԿͦΕ͓͍͍͠ͷʁͳํʢ։࠵͝ͱʹ೉қ౓͕ଟগҧ͍·͢ʣ
  94. ͋Γ͕ͱ͏͍͟͝·ͨ͠ ࢀߟจݙ: • AWS Batch – ؆୯ʹ࢖͑ͯޮ཰తͳόονίϯϐϡʔςΟϯάػೳ – AWS https://aws.amazon.com/jp/batch/

    • AWS Black Belt Online SeminarʮAWS Batchʯͷࢿྉ͓ΑͼQAެ։ http://aws.typepad.com/sajp/2017/02/aws-black-belt-online-seminar-aws- batch.html#QCPzBdn.twitter_tweet_count_m • re:Invent 2016: AWS Big Data & Machine Learning Sessionsɻ https://aws.amazon.com/blogs/big-data/reinvent-2016-aws-big-data- machine-learning-sessions/