[B! DeepLearning] tyoimariã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

tyoimari id:tyoimari

DeepLearningã«é–¢ã™ã‚‹tyoimariã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (9)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

Deep LearningåŸºç¤Žè¬›åº§æ¼”ç¿’ã‚³ãƒ³ãƒ†ãƒ³ãƒ„ å…¬é–‹ãƒšãƒ¼ã‚¸
æœ¬ãƒ—ãƒã‚°ãƒ©ãƒ ã®æœ€å¤§ã®ç‰¹å¾´ã®ä¸€ã¤ã¯ã€å…¨ã¦ã®ãƒˆãƒ”ãƒƒã‚¯ã«ã¤ã„ã¦ã€æ¼”ç¿’ã‚’ä¸å¿ƒã«æ§‹æˆã•ã‚Œã¦ã„ã‚‹ç‚¹ã§ã™ã€‚å®Ÿéš›ã«æ‰‹ã‚’å‹•ã‹ã—ãªãŒã‚‰ç†è§£ã‚’é€²ã‚ã‚‹ã“ã¨ã§ã€åŠ¹çŽ‡ã‚ˆãå¦ç¿’ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ å®Ÿéš›ã«ãƒ¢ãƒ‡ãƒ«ã‚’å¦ç¿’ã•ã›ãªãŒã‚‰æŠ€è¡“ã‚’ç¿’å¾—ã™ã‚‹æœ¬æ ¼çš„ãªæ¼”ç¿’å†…å®¹ã¨ãªã£ã¦ã„ã¾ã™ã€‚Deep Learningã¯ã€ãƒ¢ãƒ‡ãƒ«ãŒå®Ÿéš›ã«å¦ç¿’ã™ã‚‹æ§˜åã‚’è¦³æ¸¬ã—ã€ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã‚’èª¿æ•´ã™ã‚‹ã“ã¨ã§ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã«å¿œã˜ãŸãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹æœ€å¤§åŒ–ã‚’è¡Œã†ã“ã¨ãŒéžå¸¸ã«é‡è¦ãªæŠ€è¡“ã§ã™ãŒã€ã“ã®ä¸€é€£ã®æµã‚Œã‚’å…¨ã¦ã®æ¼”ç¿’ã§çµŒé¨“ã—ãªãŒã‚‰é‡è¦ãªè¦ç´ ã‚’èº«ã«ã¤ã‘ã‚‹ã“ã¨ãŒå¯èƒ½ã§ã™ã€‚
tyoimari 2018/01/24
DeepLearning

æ©Ÿæ¢°å¦ç¿’

å‹‰å¼·

ã‚ã¨ã§èªã‚€
ãƒªãƒ³ã‚¯
PYNQ-Z1ã§å§‹ã‚ã‚‹Deep Learning on FPGAå…¥é–€ï¼ˆãã®1ï¼šè³¼å…¥ã‹ã‚‰Jupyter Notebookã§Lãƒã‚«ã¾ã§ï¼‰ - Studio Ousia Engineering Blog
ã“ã‚“ã«ã¡ã¯ã€‚ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ã®æˆ¸å¡šã§ã™ã€‚ è‡ªåˆ†ã¸ã®ã‚¯ãƒªã‚¹ãƒžã‚¹ãƒ—ãƒ¬ã‚¼ãƒ³ãƒˆã¨ã—ã¦ PYNQ-Z1 ã‚’è²·ã£ãŸã®ã§ã€25ï¼…ãƒ«ãƒ¼ãƒ«ç¬¬ä¸€å¼¾ã¯ã€ŒDeep Learning on FPGAå…¥é–€ã€çš„ãªã“ã¨ã‚’ã—ã¦ã¿ãŸã„ã¨æ€ã„ã¾ã™ã€‚ ã€DISCLAIMERã€‘ã‚¹ã‚¿ãƒ¼ãƒˆæ™‚ç‚¹ã§FPGAç´ äººã§ã™ã€‚Courseraã§ã¡ã‚‡ã†ã©å¹´æœ«ã‹ã‚‰FPGAã‚³ãƒ¼ã‚¹ã‚’é–‹è¬›ã—ã¦ã„ãŸã®ã§ä½µè¡Œã—ã¦å‹‰å¼·ä¸ã§ã™ã€‚é–“é•ã£ãŸã“ã¨ã‚’æ›¸ã„ã¦ã„ãŸã‚‰å¿…è¦ã«å¿œã˜ã¦è¨‚æ£ã‚’å…¥ã‚Œã¦ã„ãã¾ã™ã€‚æœ‰è˜è€…ã®çš†æ§˜ã®ã”æŒ‡æ‘˜ã‚’æ“è¿Žã—ã¾ã™ã€‚ 1. è³¼å…¥ã‹ã‚‰èµ·å‹•ã¾ã§ å‰æ—¥èš æƒ…å ±åŽé›† PYNQ é«˜ä½åˆæˆï¼ˆHLSï¼‰ æ³¨æ–‡ èµ·å‹•ã®å‰ã« PYNQã‚·ã‚¹ãƒ†ãƒ microSDã‚«ãƒ¼ãƒ‰ã®æº–å‚™ èµ·å‹•ï¼ 2. Jupyter Notebookã«ã‚¢ã‚¯ã‚»ã‚¹ã™ã‚‹ a) LANçµŒç”±ã§ã‚¢ã‚¯ã‚»ã‚¹ PYNQã®Jupyter Notebookã®ãƒ‘ã‚¹ãƒ¯ãƒ¼ãƒ‰è¨å®š b) LANã‚±ãƒ¼ãƒ–ãƒ«ã§PCã¨ç›´æŽ¥æŽ¥ç¶š 3. Jupyter Notebo
tyoimari 2018/01/21
PYNQ

PC

deeplearning

ã‚ã¨ã§èªã‚€

ãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°

python
ãƒªãƒ³ã‚¯
æ©Ÿæ¢°å¦ç¿’ã«ä½¿ãˆã‚‹ã€ã‚ªãƒ¼ãƒ—ãƒ³ãƒ‡ãƒ¼ã‚¿ä¸€è¦§ã€€â€»éšæ™‚æ›´æ–° - Beginning AI
æ©Ÿæ¢°å¦ç¿’ã‚’ã‚„ã‚ŠãŸã„ã‚“ã ã‘ã©ã€ãƒ‡ãƒ¼ã‚¿ãŒãªã„ï¼ä»–ã®ãƒ‡ãƒ¼ã‚¿ä½¿ã£ã¦ã¿ãŸã„ï¼ ãã‚“ãªæ–¹ã®ãŸã‚ã«ã€æ©Ÿæ¢°å¦ç¿’ã«ä½¿ãˆã‚‹ã‚ªãƒ¼ãƒ—ãƒ³ãƒ‡ãƒ¼ã‚¿ã‚’é›†ã‚ã¾ã—ãŸã€‚ ä»–ã«ã‚‚ã€ã“ã®ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã‚ªã‚¹ã‚¹ãƒ¡ï¼ã¨ã„ã†ã‚‚ã®ãŒã‚ã‚Œã°ã€æ˜¯éžã”ç´¹ä»‹ã—ã¦é ‚ã‘ã‚‹ã¨å¬‰ã—ã„ã§ã™ã€‚m(__)m UC Irvine Machine Learning Repository ã‚«ãƒªãƒ•ã‚©ãƒ«ãƒ‹ã‚¢å¤§å¦ã‚¢ãƒ¼ãƒã‚¤ãƒ³æ ¡ãŒå…¬é–‹ã—ãŸã€ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã€‚351ä»¶ã®ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆãŒã‚ã‚Šå¾Œè¿°ã™ã‚‹ DATA GO ã«æ¯”ã¹ã‚Œã°å°‘ãªã„ãŒã€ã»ã¨ã‚“ã©ãŒMachine Learningç”¨ã®ãƒ‡ãƒ¼ã‚¿ãƒ»ã‚»ãƒƒãƒˆãªã®ã§ã€ã‹ãªã‚Šã‚ªã‚¹ã‚¹ãƒ¡ã€‚ UCI Machine Learning Repository ã‹ã®æœ‰åãªã‚ã‚„ã‚ã®èŠ±ï¼ˆirisï¼‰ã®ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã‚‚ã“ã“ã‹ã‚‰è¦‹ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ å›½ç«‹æƒ…å ±å¦ç ”ç©¶æ‰€ æƒ…å ±å¦ç ”ç©¶ãƒ‡ãƒ¼ã‚¿ãƒªãƒã‚¸ãƒˆãƒª ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆä¸€è¦§ yahoo,æ¥½å¤©,ãƒ‹ã‚³ãƒ‹ã‚³ãªã©ã®ãƒ‡ãƒ¼ã‚¿ãŒã‚ã‚Šã¾ã™ã€‚ DATA.GO.
tyoimari 2016/08/27
æ©Ÿæ¢°å¦ç¿’

ã‚ã¨ã§èªã‚€

deeplearning
ãƒªãƒ³ã‚¯
TensorFlow Playgroundã¯ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆã‚’ç†è§£ã™ã‚‹ã®ã«ãŠã‚¹ã‚¹ãƒ¡ - è¥¿å°¾æ³°å’Œã®ã¯ã¦ãªãƒ€ã‚¤ã‚¢ãƒªãƒ¼
ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®é‡ã¿ã‚„å„ãƒ‹ãƒ¥ãƒ¼ãƒãƒ³ãŒã©ã†ã„ã†å…¥åŠ›ã®æ™‚ã«ç™ºç«ã™ã‚‹ã®ã‹ãŒã€å¦ç¿’ã—ã¦ã„ãéŽç¨‹ã§å„æ™‚åˆ»å¯è¦–åŒ–ã•ã‚Œã¦ã¨ã¦ã‚‚è‰¯ã„æ•™æã§ã™ã€‚ http://playground.tensorflow.org/ ã†ãšã¾ãã®ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã«é–¢ã—ã¦ã€Œä¸é–“å±¤ãŒ1å±¤ã—ã‹ãªã„ã¨ã†ãšã¾ãï¼ˆç·šå½¢éžåˆ†é›¢ãªå•é¡Œï¼‰ã¯è§£ã‘ãªã„ã€ã¨ã„ã†èª¤è§£ãŒã‚ã‚‹ã‚ˆã†ãªã®ã§ã€ã¾ãšã¯1å±¤ã§ã§ãã‚‹ã¨ã„ã†çµµã‚’ç´¹ä»‹ã€‚ãªãŠåƒ•ã®ã‚¿ã‚¤ãƒ ãƒ©ã‚¤ãƒ³ä¸Šã§ã¯ id:a2c ãŒåƒ•ã‚ˆã‚Šå…ˆã«æ°—ä»˜ã„ã¦ã„ãŸã“ã¨ã‚’åèª‰ã®ãŸã‚ã«è¨€åŠã—ã¦ãŠãã¾ã™ã€‚ ã§ã€ã˜ã‚ƒã‚ã‚ˆãè¨€ã‚ã‚Œã‚‹ã€Œç·šå½¢éžåˆ†é›¢ãªå•é¡ŒãŒè§£ã‘ãªã„ã€ã£ã¦ã®ã¯ã©ã†ã„ã†ã“ã¨ã‹ã€‚ãã‚Œã¯ã“ã‚“ãªå•é¡Œè¨å®šã€‚å…¥åŠ›ã«é©å½“ãªä¿‚æ•°ã‚’æŽ›ã‘ã¦è¶³ã—åˆã‚ã›ãŸã ã‘ã§ã¯é©åˆ‡ãªå¢ƒç•Œã‚’ä½œã‚‹ã“ã¨ãŒã§ãã¾ã›ã‚“ã€‚ ã“ã†ã„ã†ã‚±ãƒ¼ã‚¹ã§ã¯ä¸é–“å±¤ã‚’è¿½åŠ ã™ã‚‹ã¨ã€ä¸é–“å±¤ãŒå…¥åŠ›ã®éžç·šå½¢ãªçµ„ã¿åˆã‚ã›ã‚’æ‹…å½“ã—ã¦ãã‚Œã‚‹ãŠã‹ã’ã§è§£ã‘ãªã‹ã£ãŸå•é¡ŒãŒè§£ã‘ã‚‹ã‚ˆã†ã«ãªã‚Šã¾ã™ã€‚ 1ã¤ç›®ã®ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã§ã¯ç‰¹å¾´é‡ã®
tyoimari 2016/04/15
TensorFlow

deepLearning

ã‚ã¨ã§èªã‚€
ãƒªãƒ³ã‚¯
ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã§æ›²ã®ã‚³ãƒ¼ãƒ‰é€²è¡Œã‚’ç”Ÿæˆã™ã‚‹ã‚¦ã‚§ãƒ–ã‚µãƒ¼ãƒ“ã‚¹ã‚’ä½œã‚Šã¾ã—ãŸ - Qiita
æ‚©ã‚ã‚‹ä½œæ›²å®¶ã®ãŸã‚ã«ã€RNN(ãƒªã‚«ãƒ¬ãƒ³ãƒˆãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯)ã‚’åˆ©ç”¨ã—ã¦ã€æ›²ã®ã‚³ãƒ¼ãƒ‰é€²è¡Œã‚’ç”Ÿæˆã™ã‚‹ã‚¦ã‚§ãƒ–ã‚µãƒ¼ãƒ“ã‚¹ Deep Chordã‚’ä½œã‚Šã¾ã—ãŸã€‚æœ¬è¨˜äº‹ã§ã¯Deep Chordã®æ¦‚è¦ã‚’ç´¹ä»‹ã—ã¾ã™ã€‚ Deep Chord Deep Chordã®RNNã®å¦ç¿’ãƒ»æŽ¨å®šã«ã¯Chainerã‚’ã€ã‚¦ã‚§ãƒ–éƒ¨åˆ†ã«ã¯Djangoã‚’åˆ©ç”¨ã—ã¦ã„ã¾ã™ã€‚ Deep Chordã®æ¦‚è¦ ã‚³ãƒ¼ãƒ‰é€²è¡Œã£ã¦ï¼Ÿ ä¾‹ãˆã°ãƒ”ã‚¢ãƒŽã®éµç›¤ã§ã€Œãƒ‰ãƒ»ãƒŸãƒ»ã‚½ã€ã¨ã‹ã€Œãƒ¬ãƒ»ãƒ•ã‚¡ãƒ»ãƒ©ã€ã¨ã‹ã€è¤‡æ•°ã®éŸ³ã‚’åŒæ™‚ã«å¼¾ãã¨ã€ãã‚Œãžã‚Œç•°ãªã‚‹éŸ¿ãã«ãªã‚Šã¾ã™ãŒã€ã“ã‚Œã‚’å’ŒéŸ³(ã‚³ãƒ¼ãƒ‰)ã¨å‘¼ã³ã¾ã™ã€‚ã“ã®å’ŒéŸ³ãŒæ§˜ã€…ã«é·ç§»ã—ã¦ã„ãã“ã¨ã§æ›²ãŒå‡ºæ¥ä¸ŠãŒã‚‹ã®ã§ã™ãŒã€ã“ã®å’ŒéŸ³ã®é·ç§»ã“ããŒã‚³ãƒ¼ãƒ‰é€²è¡Œã§ã™ã€‚ ã‚³ãƒ¼ãƒ‰é€²è¡Œã«ã¯ã€å¤šãã®æ›²ã§åˆ©ç”¨ã•ã‚Œã‚‹çŽ‹é“é€²è¡Œã¨ã„ã†ã‚‚ã®ãŒã‚ã‚Šã¾ã™ã€‚ä¾‹ãˆã°J-POPã§ã¯ã€F(ãƒ•ã‚¡ãƒ»ãƒ©ãƒ»ãƒ‰) â†’ G(ã‚½ãƒ»ã‚·ãƒ»ãƒ¬) â†’ Em(ãƒŸãƒ»ã‚½ãƒ»ã‚·) â†’ Am(ãƒ©ãƒ»ãƒ‰ãƒ»ãƒŸ)ã€
tyoimari 2016/03/12
æ©Ÿæ¢°å¦ç¿’

DeepLearning
ãƒªãƒ³ã‚¯
è‡ªç„¶è¨€èªžå‡¦ç†ã«ãŠã‘ã‚‹ç•³ã¿è¾¼ã¿ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’ç†è§£ã™ã‚‹ Â· ã‘ã‚“ã”ã®ãŠå±‹æ•·
æœ€è¿‘ã€ç•³ã¿è¾¼ã¿ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’ä½¿ã£ãŸãƒ†ã‚ã‚¹ãƒˆåˆ†é¡žã®å®Ÿé¨“ã‚’ã—ã¦ã„ã¦ã€çŸ¥è¦‹ãŒæºœã¾ã£ã¦ããŸã®ã§ãã‚Œã«ã¤ã„ã¦ä½•ã‹è¨˜äº‹ã‚’æ›¸ã“ã†ã¨æ€ã£ã¦ã„ãŸæ™‚ã«ã€ã“ã‚“ãªè¨˜äº‹ã‚’ã¿ã¤ã‘ã¾ã—ãŸã€‚ http://www.wildml.com/2015/11/understanding-convolutional-neural-networks-for-nlp ç•³ã¿è¾¼ã¿ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’è‡ªç„¶è¨€èªžå‡¦ç†ã«é©ç”¨ã™ã‚‹è©±ãªã®ã§ã™ãŒã€ã“ã®è¨˜äº‹ã€å€‹äººçš„ã«ã‚ã‹ã‚Šã‚„ã™ã„ãªã¨æ€ã£ãŸã®ã§ã€è‘—è€…ã«è¨±å¯ã‚’ã‚‚ã‚‰ã£ã¦æ—¥æœ¬èªžã«ç¿»è¨³ã—ã¾ã—ãŸã€‚ãªãŠã€ã“ã®è¨˜äº‹ã‚’èªã‚€ã«ã‚ãŸã£ã¦ã¯ã€ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã«é–¢ã™ã‚‹åŸºç¤ŽçŸ¥è˜ç¨‹åº¦ã¯å¿…è¦ã‹ã¨æ€ã‚ã‚Œã¾ã™ã€‚ â€»æ—¥æœ¬èªžã¨ã—ã¦ã‚ˆã‚Šã‚ã‹ã‚Šã‚„ã™ãè‡ªç„¶ã«ãªã‚‹ã‚ˆã†ã«ã€åŽŸæ–‡ã‚’ç›´è¨³ã—ã¦ã„ãªã„ç®‡æ‰€ã‚‚ã„ãã¤ã‹ã‚ã‚Šã¾ã™ã®ã§ã”äº†æ‰¿ãã ã•ã„ã€‚ç¿»è¨³ã®è‡´å‘½çš„ãªãƒŸã‚¹ãªã©ã‚ã‚Šã¾ã—ãŸã‚‰ã€Twitterãªã©ã§æŒ‡æ‘˜ã„ãŸã ã‘ã‚Œã°ã™ã¿ã‚„ã‹ã«ä¿®æ£ã—ã¾ã™ã€‚ ä»¥ä¸‹
tyoimari 2016/03/11
è‡ªç„¶è¨€èªžå‡¦ç†

æ©Ÿæ¢°å¦ç¿’

DeepLearning
ãƒªãƒ³ã‚¯
DeepLearningã‚’ä½¿ã£ãŸå®Ÿè£…ã‚’çºã‚ã¦ã¿ãŸ - ã®ã‚“ã³ã‚Šã—ã¦ã„ã‚‹ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ã®æ—¥è¨˜
çš†ã•ã‚“ã“ã‚“ã«ã¡ã¯ ãŠå…ƒæ°—ã§ã™ã‹ã€‚ç§ã¯å…¨ç„¶ã§ã™ã€‚ Deep Learningã€€ä¸€è¨€ã§è¨€ã†ã¨ãŸã ã®æ·±å±¤å¦ç¿’ã§ã™ãŒã€ ä½œã‚Šæ‰‹ã‚„ç”¨é€”ã«ã‚ˆã£ã¦æ§‹é€ ãŒå…¨ç„¶é•ã„ã¾ã™ã€‚ ä»Šå›žã¯é€†å¼•ãè¾žå…¸ã‚ˆã‚ã—ãã€Deep Learningã®å®Ÿè£…ã®ãƒªãƒ³ã‚¯é›†ã‚’ä½œã£ã¦ã¿ã¾ã—ãŸã€‚ ä»Šå›žã¯ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã¯å•ã‚ãšã€æŽ²è¼‰ã—ã¾ã™ã€‚ Caffeã€Theanoï¼ˆLasagneï¼‰ã€Torch7ã€Chainerãªã‚“ã§ã‚‚ã”ã–ã‚Œã§ã™ã€‚ å¾Œæ—¥ã€è¿½è¨˜ã™ã‚‹ã‹ã‚‚ãƒ»ãƒ»ãƒ» Neural Networkï¼ˆFull Connectedï¼‰ Auto Encoder Auto Encoder Denoising AutoEncoder Convolutional AutoEncoder Convolutional Neural Network Convolutional Neural Network R-CNN Fast-RCNN Faster-RCNN Recurren
tyoimari 2015/12/17
DeepLearning

ã‚ã¨ã§èªã‚€

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
Chainerã«ã‚ˆã‚‹å¤šå±¤ãƒ‘ãƒ¼ã‚»ãƒ—ãƒˆãƒãƒ³ã®å®Ÿè£… - äººå·¥çŸ¥èƒ½ã«é–¢ã™ã‚‹æ–å‰µéŒ²
ã“ã‚Œã¾ã§Deep Learningã®ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã‚’Theanoã§å®Ÿè£…ã—ã¦ããŸï¼ˆ2015/4/29ï¼‰ã‘ã‚Œã©ã€ã“ã“ã‚‰ã§å··ã§å¤§äººæ°—ã®ãƒ©ã‚¤ãƒ–ãƒ©ãƒªChainerã«ã‚‚æ‰‹ã‚’å‡ºã—ã¦ã¿ãŸã€‚Theanoã®å‹‰å¼·ã‚’å§‹ã‚ãŸã‚ã¨ã™ãã«ChainerãŒå…¬é–‹ã•ã‚ŒãŸï¼ˆ2015/6/9ï¼‰ãŒãƒ¦ãƒ¼ã‚¶ã‚„æƒ…å ±ãŒå¢—ãˆã‚‹ã¾ã§å¾…ã£ã¦ã„ãŸæ„Ÿã˜ï¼ˆç¬‘ï¼‰æœ€è¿‘ã¯ã‚³ãƒ¼ãƒ‰ã‚„å®Ÿé¨“çµæžœãªã©ã‚’å…¬é–‹ã—ã¦ãã‚Œã‚‹äººãŒå¢—ãˆã¦ããŸã®ã§éžå¸¸ã«å‚è€ƒã«ãªã£ã¦ã„ã‚‹ã€‚ç›®ã«ã¤ã„ãŸã‚‚ã®ã¯ã¦ã¶ã«ç™»éŒ²ã—ã¦ã„ã‚‹ã®ã§ã€èˆˆå‘³ã‚’æŒã£ãŸæ‰‹æ³•ã¯ãŒã—ãŒã—å‹‰å¼·ã—ã¦è¿½è©¦ã—ã¦ã„ããŸã„ã€‚ Chainerã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã¯1.3.2ã‚’ãƒ™ãƒ¼ã‚¹ã«ã—ã¦ã„ã‚‹ã€‚1.3ã‹ã‚‰PyCUDA/scikit-cudaã‚’ç‹¬è‡ªãƒ©ã‚¤ãƒ–ãƒ©ãƒªã®CuPyã«ç½®ãæ›ãˆãŸã¨ã®ã“ã¨ã§ã€ä»¥å‰ã®ã‚³ãƒ¼ãƒ‰ã¯å°‘ã—ä¿®æ£ã—ãªã„ã¨å‹•ã‹ãªã„ã‚ˆã†ã ã€‚ãã®åˆ†ã€1.3ã‹ã‚‰ã¯ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ãŒã‚·ãƒ³ãƒ—ãƒ«ã«ãªã£ã¦ã„ã¦ã¨ã¦ã‚‚ã†ã‚Œã—ã„ã€‚1.1ã®ã“ã‚ã¯ã€Chainerã¨ç›´æŽ¥é–¢ä¿‚ãªã„PyCUD
tyoimari 2015/11/30
æ©Ÿæ¢°å¦ç¿’

ã‚ã¨ã§èªã‚€

python

DeepLearning

Chainer
ãƒªãƒ³ã‚¯
Googleã€ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã‚’ã‚µãƒãƒ¼ãƒˆã—ãŸæ©Ÿæ¢°å¦ç¿’ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã€ŒTensorFlowã€ã‚’ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ã§å…¬é–‹
Googleã¯ã€è„³ã®æ´»å‹•ã‚’æ¨¡ã—ãŸãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã«ã‚ˆã£ã¦å¦ç¿’ã‚’å®Ÿç¾ã™ã‚‹ã€Œãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã€ã‚’ã‚µãƒãƒ¼ãƒˆã—ãŸæ©Ÿæ¢°å¦ç¿’ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã€ŒTensorFlowã€ã‚’ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ã§å…¬é–‹ã—ã¾ã—ãŸã€‚ãƒ©ã‚¤ã‚»ãƒ³ã‚¹ã¯Apache 2.0ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ãƒ©ã‚¤ãƒ³ã‚»ãƒ³ã‚¹ã§ã™ã€‚ Googleã¯ã™ã§ã«æ•°å¹´å‰ã‹ã‚‰ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã‚’åŒç¤¾ã®ã‚µãƒ¼ãƒ“ã‚¹ã«çµ„ã¿è¾¼ã‚“ã§ã„ã¾ã™ã€‚ã€Œç§ãŸã¡ãŒç¤¾å†…ã§ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã®åŸºç›¤ã§ã‚ã‚‹ã€ŒDistBeliefã€ã‚’é–‹ç™ºã—ãŸã®ã¯2011å¹´ã®ã“ã¨ã ã€ï¼ˆãƒ–ãƒã‚°ã€ŒTensorFlow - Googleâ€™s latest machine learning system, open sourced for everyoneã€ã‹ã‚‰ï¼‰ã€‚ TensorFlowã¯ã€ã“ã®DistBeliefã‚’ã•ã‚‰ã«å¼·åŒ–ã—ãŸã‚‚ã®ã ã¨èª¬æ˜Žã•ã‚Œã¦ã„ã¾ã™ã€‚ TensorFlow is general, flexible, porta ble, e
tyoimari 2015/11/10
æ©Ÿæ¢°å¦ç¿’

Google

ã‚ã¨ã§èªã‚€

DeepLearning
ãƒªãƒ³ã‚¯
1