{"id":3229,"date":"2026-08-05T05:46:03","date_gmt":"2026-08-04T22:46:03","guid":{"rendered":"https:\/\/blog.datacore.vn\/?p=3229"},"modified":"2026-08-07T04:17:20","modified_gmt":"2026-08-06T21:17:20","slug":"tri-tue-nhan-tao-gpt-3-fable-5","status":"publish","type":"post","link":"https:\/\/blog.datacore.vn\/vi\/tri-tue-nhan-tao-gpt-3-fable-5\/","title":{"rendered":"Tr\u00ed Tu\u1ec7 Nh\u00e2n T\u1ea1o: S\u1ef1 Th\u1eadt \u0110\u00e1ng Kinh Ng\u1ea1c T\u1eeb GPT-3 \u0110\u1ebfn Fable 5"},"content":{"rendered":"

T\u00f3m t\u1eaft:<\/strong> Tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o \u0111\u00e3 ph\u00e1t tri\u1ec3n nhanh ch\u00f3ng k\u1ec3 t\u1eeb n\u0103m 2020, t\u1eeb GPT-3 c\u1ee7a OpenAI \u0111\u1ebfn Fable 5 c\u1ee7a Anthropic v\u00e0o n\u0103m 2026. C\u00e1c b\u00e0i ki\u1ec3m tra hi\u1ec7u n\u0103ng (benchmark) li\u00ean t\u1ee5c \u0111\u01b0\u1ee3c c\u1ea3i thi\u1ec7n, nh\u01b0ng c\u00e2u h\u1ecfi th\u00fa v\u1ecb h\u01a1n kh\u00f4ng ph\u1ea3i l\u00e0 tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o \u0111\u00e3 tr\u1edf n\u00ean m\u1ea1nh m\u1ebd \u0111\u1ebfn m\u1ee9c n\u00e0o, m\u00e0 l\u00e0 lo\u1ea1i c\u00f4ng vi\u1ec7c n\u00e0o v\u1eabn c\u00f2n c\u1ea7n \u0111\u1ebfn s\u1ef1 ph\u00e1n \u0111o\u00e1n c\u1ee7a con ng\u01b0\u1eddi.<\/p>\n

N\u0103m 2020, OpenAI \u0111\u00e3 ph\u00e1t h\u00e0nh GPT-3, c\u00f9ng v\u1edbi b\u00e0i b\u00e1o \"Language models are few-shot learners\" (C\u00e1c m\u00f4 h\u00ecnh ng\u00f4n ng\u1eef l\u00e0 nh\u1eefng ng\u01b0\u1eddi h\u1ecdc \u00edt m\u1eabu), gi\u1ea3i th\u00edch chi ti\u1ebft v\u1ec1 thi\u1ebft k\u1ebf ki\u1ebfn tr\u00fac v\u00e0 c\u00e1ch qu\u00e1 tr\u00ecnh hu\u1ea5n luy\u1ec7n c\u1ee7a m\u00f4 h\u00ecnh n\u00e0y kh\u00e1c v\u1edbi c\u00e1c phi\u00ean b\u1ea3n tr\u01b0\u1edbc \u0111\u00f3.<\/p>\n

Trong khi c\u00e1c ph\u01b0\u01a1ng ph\u00e1p tr\u01b0\u1edbc \u0111\u00e2y d\u1ef1a v\u00e0o vi\u1ec7c tinh ch\u1ec9nh (fine-tuning) c\u00e1c m\u00f4 h\u00ecnh \u0111\u00e3 \u0111\u01b0\u1ee3c hu\u1ea5n luy\u1ec7n s\u1eb5n tr\u00ean c\u00e1c t\u1eadp d\u1eef li\u1ec7u chuy\u00ean bi\u1ec7t cho t\u1eebng nhi\u1ec7m v\u1ee5, ch\u1ee9a s\u1ed1 l\u01b0\u1ee3ng l\u1edbn c\u00e1c v\u00ed d\u1ee5 \u0111\u01b0\u1ee3c g\u00e1n nh\u00e3n, th\u00ec GPT-3 l\u1ea1i \u0111i theo m\u1ed9t h\u01b0\u1edbng kh\u00e1c: m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c m\u1edf r\u1ed9ng quy m\u00f4 l\u00ean 175 t\u1ef7 tham s\u1ed1 v\u00e0 \u00e1p d\u1ee5ng cho c\u00e1c t\u00e1c v\u1ee5 ti\u1ebfp theo m\u00e0 kh\u00f4ng c\u1ea7n b\u1ea5t k\u1ef3 c\u1eadp nh\u1eadt gradient n\u00e0o, nh\u01b0ng v\u1eabn \u0111\u1ea1t hi\u1ec7u su\u1ea5t cao tr\u00ean h\u1ea7u h\u1ebft c\u00e1c b\u00e0i ki\u1ec3m tra NLP.<\/p>\n

\u0110\u00e1ng ch\u00fa \u00fd, c\u00e1c t\u00e1c gi\u1ea3 b\u00e1o c\u00e1o r\u1eb1ng nh\u1eefng ng\u01b0\u1eddi \u0111\u00e1nh gi\u00e1 g\u1eb7p kh\u00f3 kh\u0103n trong vi\u1ec7c ph\u00e2n bi\u1ec7t c\u00e1c b\u00e0i b\u00e1o tin t\u1ee9c do GPT-3 t\u1ea1o ra v\u1edbi nh\u1eefng b\u00e0i do con ng\u01b0\u1eddi vi\u1ebft.<\/p>\n

T\u1eeb GPT-3 \u0111\u1ebfn Fable 5: m\u1ecdi chuy\u1ec7n b\u1eaft \u0111\u1ea7u nh\u01b0 th\u1ebf n\u00e0o?<\/h2>\n
\"Tr\u00ed<\/figure>\n

V\u1ec1 m\u1eb7t ph\u01b0\u01a1ng ph\u00e1p, c\u00e1c t\u00e1c gi\u1ea3 \u0111\u00e3 \u00e1p d\u1ee5ng ki\u1ebfn tr\u00fac transformer v\u00e0 hu\u1ea5n luy\u1ec7n m\u1ed9t lo\u1ea1t m\u00f4 h\u00ecnh c\u00f3 quy m\u00f4 t\u1eeb 125 tri\u1ec7u \u0111\u1ebfn 175 t\u1ef7 tham s\u1ed1 nh\u1eb1m nghi\u00ean c\u1ee9u c\u00e1ch hi\u1ec7u n\u0103ng c\u1ee7a tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o thay \u0111\u1ed5i theo k\u00edch th\u01b0\u1edbc m\u00f4 h\u00ecnh. Vi\u1ec7c \u0111\u00e1nh gi\u00e1 \u0111\u01b0\u1ee3c th\u1ef1c hi\u1ec7n theo ba ch\u1ebf \u0111\u1ed9: few-shot (m\u1ed9t s\u1ed1 \u00edt m\u1eabu), trong \u0111\u00f3 m\u00f4 h\u00ecnh nh\u1eadn \u0111\u01b0\u1ee3c m\u00f4 t\u1ea3 nhi\u1ec7m v\u1ee5 c\u00f9ng v\u1edbi m\u1ed9t v\u00e0i minh h\u1ecda; one-shot (m\u1ed9t m\u1eabu), ch\u1ec9 c\u00f3 m\u1ed9t minh h\u1ecda duy nh\u1ea5t \u0111\u01b0\u1ee3c \u0111\u01b0a ra; v\u00e0 zero-shot (kh\u00f4ng m\u1eabu), trong \u0111\u00f3 m\u00f4 h\u00ecnh ch\u1ec9 nh\u1eadn \u0111\u01b0\u1ee3c m\u1ed9t h\u01b0\u1edbng d\u1eabn b\u1eb1ng ng\u00f4n ng\u1eef t\u1ef1 nhi\u00ean.<\/p>\n

Nh\u1edd th\u00e0nh c\u00f4ng n\u00e0y, GPT-3 \u0111\u00e3 tr\u1edf th\u00e0nh m\u1ed9t m\u00f4 h\u00ecnh t\u1ea1o sinh tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o \u0111\u01b0\u1ee3c c\u00f4ng nh\u1eadn r\u1ed9ng r\u00e3i v\u00e0 \u0111\u1ea1t th\u00e0nh c\u00f4ng th\u01b0\u01a1ng m\u1ea1i \u0111\u00e1ng k\u1ec3.<\/p>\n

Ti\u1ebfn \u0111\u1ebfn n\u0103m 2026. Anthropic sau \u0111\u00f3 \u0111\u00e3 ph\u00e1t h\u00e0nh Claude Opus 5<\/a>, v\u00e0 gi\u1edd \u0111\u00e2y l\u00e0 Fable 5, m\u1ed9t trong nh\u1eefng m\u00f4 h\u00ecnh c\u00f3 n\u0103ng l\u1ef1c cao nh\u1ea5t m\u00e0 c\u00f4ng ch\u00fang t\u1eebng \u0111\u01b0\u1ee3c ti\u1ebfp c\u1eadn. Anthropic ch\u01b0a c\u00f4ng b\u1ed1 l\u00fd thuy\u1ebft hay ki\u1ebfn tr\u00fac \u0111\u1eb1ng sau m\u00f4 h\u00ecnh n\u00e0y, nh\u01b0ng c\u00f3 th\u1ec3 h\u1ee3p l\u00fd khi cho r\u1eb1ng Fable 5 \u0111\u01b0\u1ee3c x\u00e2y d\u1ef1ng d\u1ef1a tr\u00ean m\u1ed9t s\u1ed1 h\u01b0\u1edbng \u0111i m\u00e0 OpenAI \u0111\u00e3 g\u1ee3i \u00fd v\u00e0o n\u0103m 2020.<\/p>\n

Tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o ng\u00e0y nay m\u1ea1nh \u0111\u1ebfn m\u1ee9c n\u00e0o?<\/h2>\n

Tr\u00ean c\u00e1c t\u00e1c v\u1ee5 \u0111\u00f2i h\u1ecfi kh\u1ea3 n\u0103ng suy lu\u1eadn ph\u1ee9c t\u1ea1p, t\u1eeb gi\u1ea3i c\u00e1c b\u00e0i to\u00e1n cho \u0111\u1ebfn t\u00e1i t\u1ea1o m\u00e3 ngu\u1ed3n c\u1ee7a m\u1ed9t \u1ee9ng d\u1ee5ng di \u0111\u1ed9ng ch\u1ec9 t\u1eeb m\u1ed9t \u1ea3nh ch\u1ee5p m\u00e0n h\u00ecnh, Fable 5 \u0111\u1ea1t k\u1ebft qu\u1ea3 ngang b\u1eb1ng ho\u1eb7c v\u01b0\u1ee3t qua c\u00e1c b\u00e0i ki\u1ec3m tra hi\u1ec7u n\u0103ng m\u1ea1nh nh\u1ea5t t\u1eebng \u0111\u01b0\u1ee3c c\u00f4ng b\u1ed1.<\/p>\n

\"Giao<\/figure>\n

Gi\u1ed1ng nh\u01b0 GPT-3 t\u1eebng l\u00e0m n\u1ea3y sinh h\u00e0ng lo\u1ea1t v\u1ea5n \u0111\u1ec1 x\u00e3 h\u1ed9i v\u00e0 th\u01b0\u01a1ng m\u1ea1i, d\u1ec5 th\u1ea5y nh\u1ea5t l\u00e0 vi\u1ec7c m\u1ed9t \u0111\u1ed9c gi\u1ea3 b\u00ecnh th\u01b0\u1eddng ph\u1ea3i ph\u00e2n bi\u1ec7t m\u1ed9t b\u00e0i b\u00e1o h\u1ee3p l\u1ec7 v\u1edbi m\u1ed9t b\u00e0i b\u00e1o t\u1ed5ng h\u1ee3p v\u00e0 c\u00f3 th\u1ec3 g\u00e2y h\u1ea1i nh\u01b0 th\u1ebf n\u00e0o, Fable 5 c\u0169ng \u0111\u1eb7t ra m\u1ed9t v\u1ea5n \u0111\u1ec1 v\u1eeba mang t\u00ednh tri\u1ebft h\u1ecdc v\u1eeba mang t\u00ednh k\u1ef9 thu\u1eadt: con ng\u01b0\u1eddi c\u00f2n l\u1ea1i c\u00f4ng vi\u1ec7c g\u00ec \u0111\u1ec3 l\u00e0m, v\u00e0 ch\u00fang ta th\u1ef1c s\u1ef1 hi\u1ec3u tr\u00ed tu\u1ec7 l\u00e0 g\u00ec?<\/p>\n

S\u1ef1 thay \u0111\u1ed5i n\u00e0y \u0111\u00e3 th\u1ec3 hi\u1ec7n r\u00f5 qua c\u00e1ch c\u00e1c doanh nghi\u1ec7p l\u00ean k\u1ebf ho\u1ea1ch tuy\u1ec3n d\u1ee5ng xoay quanh t\u00e1c nh\u00e2n AI<\/a>.<\/p>\n

Con ng\u01b0\u1eddi c\u00f2n l\u1ea1i nh\u1eefng vi\u1ec7c g\u00ec \u0111\u1ec3 l\u00e0m?<\/h2>\n

Tuy nhi\u00ean, ph\u1ea7n l\u1edbn c\u00f4ng vi\u1ec7c th\u1ef1c t\u1ebf kh\u00f4ng n\u1eb1m \u1edf vi\u1ec7c gi\u1ea3i c\u00e1c b\u00e0i to\u00e1n \u0111\u00e3 \u0111\u01b0\u1ee3c \u0111\u1ecbnh ngh\u0129a r\u00f5 r\u00e0ng. C\u00f4ng vi\u1ec7c \u0111\u00f2i h\u1ecfi ph\u00e1n \u0111o\u00e1n, ch\u1ee9 kh\u00f4ng ph\u1ea3i ghi nh\u1edb, v\u1eabn thu\u1ed9c v\u1ec1 l\u00e3nh \u0111\u1ecba ri\u00eang c\u1ee7a con ng\u01b0\u1eddi.<\/p>\n

H\u00e3y xem x\u00e9t c\u00f4ng vi\u1ec7c t\u00ecm hi\u1ec3u m\u1ed9t th\u1ecb tr\u01b0\u1eddng: x\u00e1c \u0111\u1ecbnh kh\u00e1ch h\u00e0ng ti\u1ec1m n\u0103ng th\u1ef1c s\u1ef1 l\u00e0 ai, ph\u00e2n bi\u1ec7t \u0111i\u1ec1u h\u1ecd mu\u1ed1n v\u1edbi \u0111i\u1ec1u h\u1ecd th\u1ef1c s\u1ef1 c\u1ea7n, r\u1ed3i t\u1eeb \u0111\u00f3 x\u00e2y d\u1ef1ng m\u1ed9t gi\u1ea3i ph\u00e1p gi\u1ea3i quy\u1ebft \u0111\u00fang v\u1ea5n \u0111\u1ec1 c\u1ed1t l\u00f5i thay v\u00ec v\u1ea5n \u0111\u1ec1 \u0111\u01b0\u1ee3c n\u00eau ra b\u1ec1 m\u1eb7t. \u0110\u00f3 l\u00e0 c\u00f4ng vi\u1ec7c m\u00e0 hi\u1ec7n t\u1ea1i kh\u00f4ng m\u00f4 h\u00ecnh n\u00e0o c\u00f3 th\u1ec3 t\u1ef1 m\u00ecnh th\u1ef1c hi\u1ec7n \u0111\u01b0\u1ee3c.<\/p>\n

C\u00e2u h\u1ecfi th\u01b0\u1eddng g\u1eb7p<\/h2>\n

GPT-3 l\u00e0 g\u00ec?<\/strong>
GPT-3 l\u00e0 m\u00f4 h\u00ecnh ng\u00f4n ng\u1eef v\u1edbi 175 t\u1ef7 tham s\u1ed1 m\u00e0 OpenAI gi\u1edbi thi\u1ec7u v\u00e0o n\u0103m 2020 trong b\u00e0i b\u00e1o \"Language models are few-shot learners\", \u0111\u01b0\u1ee3c c\u00f4ng b\u1ed1 trong Advances in Neural Information Processing Systems 33 (2020): 1877-1901.<\/p>\n

Fable 5 l\u00e0 g\u00ec?<\/strong>
Fable 5 l\u00e0 m\u1ed9t m\u00f4 h\u00ecnh ng\u00f4n ng\u1eef l\u1edbn do Anthropic ph\u00e1t h\u00e0nh n\u0103m 2026, \u0111\u01b0\u1ee3c c\u00f4ng b\u1ed1 c\u00f9ng v\u1edbi Claude Mythos 5, \u0111\u1ea1t k\u1ebft qu\u1ea3 ngang b\u1eb1ng ho\u1eb7c v\u01b0\u1ee3t qua c\u00e1c b\u00e0i ki\u1ec3m tra hi\u1ec7u n\u0103ng m\u1ea1nh nh\u1ea5t t\u1eebng \u0111\u01b0\u1ee3c c\u00f4ng b\u1ed1 \u0111\u1ed1i v\u1edbi c\u00e1c t\u00e1c v\u1ee5 suy lu\u1eadn ph\u1ee9c t\u1ea1p.<\/p>\n

Tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o c\u00f3 thay th\u1ebf \u0111\u01b0\u1ee3c ph\u00e1n \u0111o\u00e1n c\u1ee7a con ng\u01b0\u1eddi kh\u00f4ng?<\/strong>
Kh\u00f4ng. Tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o hi\u1ec7n \u0111\u1ea1i v\u1eabn ch\u01b0a th\u1ec3 thay th\u1ebf con ng\u01b0\u1eddi trong nh\u1eefng c\u00f4ng vi\u1ec7c \u0111\u00f2i h\u1ecfi ph\u00e1n \u0111o\u00e1n, ch\u1eb3ng h\u1ea1n nh\u01b0 t\u00ecm hi\u1ec3u m\u1ed9t th\u1ecb tr\u01b0\u1eddng ho\u1eb7c ph\u00e2n bi\u1ec7t \u0111i\u1ec1u kh\u00e1ch h\u00e0ng mu\u1ed1n v\u1edbi \u0111i\u1ec1u h\u1ecd th\u1ef1c s\u1ef1 c\u1ea7n, v\u1eabn l\u00e0 l\u00e3nh \u0111\u1ecba ri\u00eang c\u1ee7a con ng\u01b0\u1eddi m\u00e0 hi\u1ec7n t\u1ea1i kh\u00f4ng m\u00f4 h\u00ecnh n\u00e0o c\u00f3 th\u1ec3 t\u1ef1 m\u00ecnh th\u1ef1c hi\u1ec7n \u0111\u01b0\u1ee3c.<\/p>\n

Ngu\u1ed3n:<\/p>\n

Th\u00f4ng b\u00e1o c\u1ee7a Anthropic: \"Claude Fable 5 v\u00e0 Claude Mythos 5\": https:\/\/www.anthropic.com\/news\/claude-fable-5-mythos-5<\/a><\/p>\n

Brown, Tom, et al. \"Language models are few-shot learners.\" Advances in neural information processing systems 33 (2020): 1877-1901.<\/p>\n

T\u00f2 m\u00f2 c\u00e1c m\u00f4 h\u00ecnh AI m\u1edbi nh\u1ea5t so s\u00e1nh v\u1edbi nhau ra sao? Xem B\u1ea3ng X\u1ebfp H\u1ea1ng AI<\/a> c\u1ee7a DataCore, b\u1ea3ng x\u1ebfp h\u1ea1ng th\u1eddi gian th\u1ef1c mi\u1ec5n ph\u00ed cho c\u00e1c m\u00f4 h\u00ecnh AI h\u00e0ng \u0111\u1ea7u, c\u1eadp nh\u1eadt li\u00ean t\u1ee5c.<\/p>\n

Mu\u1ed1n xem b\u1ea3ng x\u1ebfp h\u1ea1ng m\u00f4 h\u00ecnh AI c\u1eadp nh\u1eadt theo th\u1eddi gian th\u1ef1c?<\/strong> Xem B\u1ea3ng X\u1ebfp H\u1ea1ng AI c\u1ee7a DataCore<\/a>.<\/p>","protected":false},"excerpt":{"rendered":"

Tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o \u0111\u00e3 \u0111i t\u1eeb GPT-3 n\u0103m 2020 \u0111\u1ebfn Fable 5 c\u1ee7a Anthropic n\u0103m 2026. \u0110\u00e2y l\u00e0 c\u00e1ch quy m\u00f4 m\u00f4 h\u00ecnh thay \u0111\u1ed5i, c\u00e1c b\u00e0i ki\u1ec3m tra hi\u1ec7u n\u0103ng cho th\u1ea5y g\u00ec, v\u00e0 \u0111i\u1ec1u g\u00ec v\u1eabn c\u1ea7n s\u1ef1 ph\u00e1n \u0111o\u00e1n c\u1ee7a con ng\u01b0\u1eddi.<\/p>","protected":false},"author":19,"featured_media":3194,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"content-type":"","_uag_custom_page_level_css":"","_swt_meta_header_display":false,"_swt_meta_footer_display":false,"_swt_meta_site_title_display":false,"_swt_meta_sticky_header":false,"_swt_meta_transparent_header":false,"footnotes":""},"categories":[19],"tags":[2426,2437,1285],"class_list":["post-3229","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-blog-vi","tag-dc-2026-w32-vi-vi","tag-du-lieu-ai-vi","tag-tri-tue-nhan-tao-vi"],"uagb_featured_image_src":{"full":["https:\/\/blog.datacore.vn\/wp-content\/uploads\/2026\/08\/artificial-neural-network-chip.jpg",1280,1024,false],"thumbnail":["https:\/\/blog.datacore.vn\/wp-content\/uploads\/2026\/08\/artificial-neural-network-chip-150x150.jpg",150,150,true],"medium":["https:\/\/blog.datacore.vn\/wp-content\/uploads\/2026\/08\/artificial-neural-network-chip-300x240.jpg",300,240,true],"medium_large":["https:\/\/blog.datacore.vn\/wp-content\/uploads\/2026\/08\/artificial-neural-network-chip-768x614.jpg",768,614,true],"large":["https:\/\/blog.datacore.vn\/wp-content\/uploads\/2026\/08\/artificial-neural-network-chip-1024x819.jpg",1024,819,true],"1536x1536":["https:\/\/blog.datacore.vn\/wp-content\/uploads\/2026\/08\/artificial-neural-network-chip.jpg",1280,1024,false],"2048x2048":["https:\/\/blog.datacore.vn\/wp-content\/uploads\/2026\/08\/artificial-neural-network-chip.jpg",1280,1024,false],"trp-custom-language-flag":["https:\/\/blog.datacore.vn\/wp-content\/uploads\/2026\/08\/artificial-neural-network-chip-15x12.jpg",15,12,true]},"uagb_author_info":{"display_name":"DataCore Marketing","author_link":"https:\/\/blog.datacore.vn\/vi\/author\/datacore_marketing\/"},"uagb_comment_info":0,"uagb_excerpt":"Tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o \u0111\u00e3 \u0111i t\u1eeb GPT-3 n\u0103m 2020 \u0111\u1ebfn Fable 5 c\u1ee7a Anthropic n\u0103m 2026. \u0110\u00e2y l\u00e0 c\u00e1ch quy m\u00f4 m\u00f4 h\u00ecnh thay \u0111\u1ed5i, c\u00e1c b\u00e0i ki\u1ec3m tra hi\u1ec7u n\u0103ng cho th\u1ea5y g\u00ec, v\u00e0 \u0111i\u1ec1u g\u00ec v\u1eabn c\u1ea7n s\u1ef1 ph\u00e1n \u0111o\u00e1n c\u1ee7a con ng\u01b0\u1eddi.","_links":{"self":[{"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/posts\/3229","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/users\/19"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/comments?post=3229"}],"version-history":[{"count":4,"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/posts\/3229\/revisions"}],"predecessor-version":[{"id":3502,"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/posts\/3229\/revisions\/3502"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/media\/3194"}],"wp:attachment":[{"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/media?parent=3229"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/categories?post=3229"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.datacore.vn\/vi\/wp-json\/wp\/v2\/tags?post=3229"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}