{"id":18435,"date":"2024-04-24T15:48:05","date_gmt":"2024-04-24T08:48:05","guid":{"rendered":"https:\/\/gcloudvn.com\/?p=18435"},"modified":"2024-04-25T11:24:19","modified_gmt":"2024-04-25T04:24:19","slug":"introducing-llm-fine-tuning-and-evaluation-in-bigquery","status":"publish","type":"post","link":"https:\/\/gcloudvn.com\/en\/kienthuc\/introducing-llm-fine-tuning-and-evaluation-in-bigquery\/","title":{"rendered":"Introducing LLM fine-tuning and evaluation in BigQuery"},"content":{"rendered":"<section class=\"wpb-content-wrapper\"><div class=\"vc_row wpb_row vc_row-fluid\"><div class=\"wpb_column vc_column_container vc_col-sm-12\"><div class=\"vc_column-inner\"><div class=\"wpb_wrapper\">\n\t<div class=\"wpb_text_column wpb_content_element \" >\n\t\t<div class=\"wpb_wrapper\">\n\t\t\t<p><span style=\"font-weight: 400;\">BigQuery cho ph\u00e9p b\u1ea1n ph\u00e2n t\u00edch d\u1eef li\u1ec7u c\u1ee7a m\u00ecnh b\u1eb1ng c\u00e1ch s\u1eed d\u1ee5ng m\u1ed9t lo\u1ea1t c\u00e1c m\u00f4 h\u00ecnh ng\u00f4n ng\u1eef l\u1edbn (LLMs) \u0111\u01b0\u1ee3c l\u01b0u tr\u1eef trong Vertex AI bao g\u1ed3m Gemini 1.0 Pro, Gemini 1.0 Pro Vision v\u00e0 text-bison. C\u00e1c m\u00f4 h\u00ecnh n\u00e0y ho\u1ea1t \u0111\u1ed9ng t\u1ed1t cho nhi\u1ec1u nhi\u1ec7m v\u1ee5 nh\u01b0 t\u00f3m t\u1eaft v\u0103n b\u1ea3n, ph\u00e2n t\u00edch t\u00e2m tr\u1ea1ng, vv. ch\u1ec9 b\u1eb1ng c\u00e1ch s\u1eed d\u1ee5ng k\u1ef9 thu\u1eadt prompt. Tuy nhi\u00ean, trong m\u1ed9t s\u1ed1 t\u00ecnh hu\u1ed1ng, vi\u1ec7c tinh ch\u1ec9nh b\u1ed5 sung th\u00f4ng qua vi\u1ec7c tinh ch\u1ec9nh m\u00f4 h\u00ecnh l\u00e0 c\u1ea7n thi\u1ebft, nh\u01b0 khi h\u00e0nh vi mong \u0111\u1ee3i c\u1ee7a m\u00f4 h\u00ecnh kh\u00f3 m\u00f4 t\u1ea3 m\u1ed9t c\u00e1ch ng\u1eafn g\u1ecdn trong m\u1ed9t prompt, ho\u1eb7c khi prompt kh\u00f4ng t\u1ea1o ra k\u1ebft qu\u1ea3 mong \u0111\u1ee3i m\u1ed9t c\u00e1ch \u0111\u1ed3ng nh\u1ea5t \u0111\u1ee7.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Vi\u1ec7c tinh ch\u1ec9nh b\u1ed5 sung c\u0169ng gi\u00fap m\u00f4 h\u00ecnh h\u1ecdc c\u00e1c phong c\u00e1ch ph\u1ea3n \u1ee9ng c\u1ee5 th\u1ec3 (v\u00ed d\u1ee5: ng\u1eafn g\u1ecdn ho\u1eb7c d\u00e0i d\u00f2ng), c\u00e1c h\u00e0nh vi m\u1edbi (v\u00ed d\u1ee5: tr\u1ea3 l\u1eddi d\u01b0\u1edbi d\u1ea1ng m\u1ed9t nh\u00e2n v\u1eadt c\u1ee5 th\u1ec3), ho\u1eb7c c\u1eadp nh\u1eadt m\u00ecnh v\u1edbi th\u00f4ng tin m\u1edbi.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">H\u00f4m nay, Google th\u00f4ng b\u00e1o v\u1ec1 vi\u1ec7c h\u1ed7 tr\u1ee3 tinh ch\u1ec9nh t\u00f9y ch\u1ec9nh cho LLM trong BigQuery th\u00f4ng qua vi\u1ec7c tinh ch\u1ec9nh gi\u00e1m s\u00e1t. Tinh ch\u1ec9nh gi\u00e1m s\u00e1t th\u00f4ng qua BigQuery s\u1eed d\u1ee5ng m\u1ed9t t\u1eadp d\u1eef li\u1ec7u c\u00f3 c\u00e1c v\u00ed d\u1ee5 v\u1ec1 v\u0103n b\u1ea3n \u0111\u1ea7u v\u00e0o (prompt) v\u00e0 v\u0103n b\u1ea3n \u0111\u1ea7u ra l\u00fd t\u01b0\u1edfng mong \u0111\u1ee3i (nh\u00e3n), v\u00e0 tinh ch\u1ec9nh m\u00f4 h\u00ecnh \u0111\u1ec3 m\u00f4 ph\u1ecfng h\u00e0nh vi ho\u1eb7c nhi\u1ec7m v\u1ee5 \u0111\u01b0\u1ee3c ng\u1ee5 \u00fd t\u1eeb nh\u1eefng v\u00ed d\u1ee5 n\u00e0y. H\u00e3y xem c\u00e1ch th\u1ee9c ho\u1ea1t \u0111\u1ed9ng n\u00e0y.<\/span><\/p>\n<div id=\"ez-toc-container\" class=\"ez-toc-v2_0_83 counter-hierarchy ez-toc-counter ez-toc-grey ez-toc-container-direction\">\n<div class=\"ez-toc-title-container\">\n<p class=\"ez-toc-title\" style=\"cursor:inherit\">M\u1ee5c L\u1ee5c<\/p>\n<span class=\"ez-toc-title-toggle\"><a href=\"#\" class=\"ez-toc-pull-right ez-toc-btn ez-toc-btn-xs ez-toc-btn-default ez-toc-toggle\" aria-label=\"Toggle Table of Content\"><span class=\"ez-toc-js-icon-con\"><span class=\"\"><span class=\"eztoc-hide\" style=\"display:none;\">Toggle<\/span><span class=\"ez-toc-icon-toggle-span\"><svg style=\"fill: #999;color:#999\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" class=\"list-377408\" width=\"20px\" height=\"20px\" viewBox=\"0 0 24 24\" fill=\"none\"><path d=\"M6 6H4v2h2V6zm14 0H8v2h12V6zM4 11h2v2H4v-2zm16 0H8v2h12v-2zM4 16h2v2H4v-2zm16 0H8v2h12v-2z\" fill=\"currentColor\"><\/path><\/svg><svg style=\"fill: #999;color:#999\" class=\"arrow-unsorted-368013\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"10px\" height=\"10px\" viewBox=\"0 0 24 24\" version=\"1.2\" baseProfile=\"tiny\"><path d=\"M18.2 9.3l-6.2-6.3-6.2 6.3c-.2.2-.3.4-.3.7s.1.5.3.7c.2.2.4.3.7.3h11c.3 0 .5-.1.7-.3.2-.2.3-.5.3-.7s-.1-.5-.3-.7zM5.8 14.7l6.2 6.3 6.2-6.3c.2-.2.3-.5.3-.7s-.1-.5-.3-.7c-.2-.2-.4-.3-.7-.3h-11c-.3 0-.5.1-.7.3-.2.2-.3.5-.3.7s.1.5.3.7z\"\/><\/svg><\/span><\/span><\/span><\/a><\/span><\/div>\n<nav><ul class='ez-toc-list ez-toc-list-level-1 ' ><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-1\" href=\"https:\/\/gcloudvn.com\/en\/kienthuc\/introducing-llm-fine-tuning-and-evaluation-in-bigquery\/#Cac_tinh_nang_tieu_bieu\" >C\u00e1c t\u00ednh n\u0103ng ti\u00eau bi\u1ec3u<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-2\" href=\"https:\/\/gcloudvn.com\/en\/kienthuc\/introducing-llm-fine-tuning-and-evaluation-in-bigquery\/#Dataset\" >Dataset\u00a0<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-3\" href=\"https:\/\/gcloudvn.com\/en\/kienthuc\/introducing-llm-fine-tuning-and-evaluation-in-bigquery\/#Hieu_ve_hieu_suat_co_ban_cua_mo_hinh_text-bison\" >Hi\u1ec3u v\u1ec1 hi\u1ec7u su\u1ea5t c\u01a1 b\u1ea3n c\u1ee7a m\u00f4 h\u00ecnh text-bison<\/a><\/li><\/ul><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-4\" href=\"https:\/\/gcloudvn.com\/en\/kienthuc\/introducing-llm-fine-tuning-and-evaluation-in-bigquery\/#Tao_mot_mo_hinh_duoc_tinh_chinh\" >T\u1ea1o m\u1ed9t m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-5\" href=\"https:\/\/gcloudvn.com\/en\/kienthuc\/introducing-llm-fine-tuning-and-evaluation-in-bigquery\/#Danh_gia_hieu_xuat_tinh_chinh_model\" >\u0110\u00e1nh gi\u00e1 hi\u1ec7u xu\u1ea5t tinh ch\u1ec9nh model<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-6\" href=\"https:\/\/gcloudvn.com\/en\/kienthuc\/introducing-llm-fine-tuning-and-evaluation-in-bigquery\/#Danh_gia_dua_tren_chi_so_cho_mo_hinh_duoc_tinh_chinh\" >\u0110\u00e1nh gi\u00e1 d\u1ef1a tr\u00ean ch\u1ec9 s\u1ed1 cho m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh<\/a><\/li><\/ul><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-7\" href=\"https:\/\/gcloudvn.com\/en\/kienthuc\/introducing-llm-fine-tuning-and-evaluation-in-bigquery\/#San_sang_cho_suy_luan\" >S\u1eb5n s\u00e0ng cho suy lu\u1eadn\u00a0<\/a><\/li><\/ul><\/nav><\/div>\n<h2><span class=\"ez-toc-section\" id=\"Cac_tinh_nang_tieu_bieu\"><\/span><b>C\u00e1c t\u00ednh n\u0103ng ti\u00eau bi\u1ec3u<\/b><span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p><span style=\"font-weight: 400;\">\u0110\u1ec3 minh h\u1ecda vi\u1ec7c tinh ch\u1ec9nh m\u00f4 h\u00ecnh, h\u00e3y xem x\u00e9t m\u1ed9t v\u1ea5n \u0111\u1ec1 ph\u00e2n lo\u1ea1i s\u1eed d\u1ee5ng d\u1eef li\u1ec7u v\u0103n b\u1ea3n. Ch\u00fang ta s\u1ebd s\u1eed d\u1ee5ng m\u1ed9t t\u1eadp d\u1eef li\u1ec7u chuy\u1ec3n \u0111\u1ed5i y t\u1ebf v\u00e0 y\u00eau c\u1ea7u m\u00f4 h\u00ecnh c\u1ee7a ch\u00fang ta s\u1ebd l\u00e0 ph\u00e2n lo\u1ea1i m\u1ed9t b\u1ea3n ghi nh\u1ea5t \u0111\u1ecbnh v\u00e0o m\u1ed9t trong 17 danh m\u1ee5c, v\u00ed d\u1ee5 nh\u01b0 &#8216;D\u1ecb \u1ee9ng\/ Mi\u1ec5n d\u1ecbch&#8217;, &#8216;Nha khoa&#8217;, &#8216;Tim m\u1ea1ch\/ Ph\u1ed5i&#8217;, vv.<\/span><\/p>\n<h3><span class=\"ez-toc-section\" id=\"Dataset\"><\/span><b>Dataset\u00a0<\/b><span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p><span style=\"font-weight: 400;\">T\u1eadp d\u1eef li\u1ec7u c\u1ee7a ch\u00fang ta \u0111\u01b0\u1ee3c<\/span><span style=\"font-weight: 400;\">\u00a0cung c\u1ea5p tr\u00ean <\/span><span style=\"font-weight: 400;\">Kaggle<\/span><span style=\"font-weight: 400;\">. \u0110\u1ec3 tinh ch\u1ec9nh v\u00e0 \u0111\u00e1nh gi\u00e1 m\u00f4 h\u00ecnh n\u00e0y, tr\u01b0\u1edbc ti\u00ean ch\u00fang ta s\u1ebd t\u1ea1o m\u1ed9t b\u1ea3ng \u0111\u00e1nh gi\u00e1 v\u00e0 m\u1ed9t b\u1ea3ng hu\u1ea5n luy\u1ec7n trong BigQuery b\u1eb1ng c\u00e1ch s\u1eed d\u1ee5ng m\u1ed9t ph\u1ea7n c\u1ee7a d\u1eef li\u1ec7u n\u00e0y c\u00f3 s\u1eb5n trong Cloud Storage nh\u01b0 sau:<\/span><\/p>\n<p>&#8212; Create a eval table<\/p>\n<p>LOAD DATA INTO<br \/>\nbqml_tutorial.medical_transcript_eval<br \/>\nFROM FILES( format=&#8217;NEWLINE_DELIMITED_JSON&#8217;,<br \/>\nuris = [&#8216;gs:\/\/cloud-samples-data\/vertex-ai\/model-evaluation\/peft_eval_sample.jsonl&#8217;] )<\/p>\n<p>&#8212; Create a train table<\/p>\n<p>LOAD DATA INTO<br \/>\nbqml_tutorial.medical_transcript_train<br \/>\nFROM FILES( format=&#8217;NEWLINE_DELIMITED_JSON&#8217;,<br \/>\nuris = [&#8216;gs:\/\/cloud-samples-data\/vertex-ai\/model-evaluation\/peft_train_sample.jsonl&#8217;] )<\/p>\n<p><span style=\"font-weight: 400;\">B\u1ed9 d\u1eef li\u1ec7u hu\u1ea5n luy\u1ec7n v\u00e0 \u0111\u00e1nh gi\u00e1 c\u00f3 m\u1ed9t c\u1ed9t &#8216;input_text&#8217; ch\u1ee9a b\u1ea3n ghi, v\u00e0 m\u1ed9t c\u1ed9t &#8216;output_text&#8217; ch\u1ee9a nh\u00e3n ho\u1eb7c Ground truth.<\/span><\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-18429 size-full\" src=\"https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry.jpg\" alt=\"\" width=\"601\" height=\"388\" srcset=\"https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry.jpg 601w, https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry-18x12.jpg 18w\" sizes=\"auto, (max-width: 601px) 100vw, 601px\" \/><\/p>\n<h3><span class=\"ez-toc-section\" id=\"Hieu_ve_hieu_suat_co_ban_cua_mo_hinh_text-bison\"><\/span><b>Hi\u1ec3u v\u1ec1 hi\u1ec7u su\u1ea5t c\u01a1 b\u1ea3n c\u1ee7a m\u00f4 h\u00ecnh text-bison<\/b><span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p><span style=\"font-weight: 400;\">Tr\u01b0\u1edbc h\u1ebft, h\u00e3y x\u00e1c \u0111\u1ecbnh m\u1ed9t \u0111i\u1ec3m c\u01a1 b\u1ea3n v\u1ec1 hi\u1ec7u su\u1ea5t cho m\u00f4 h\u00ecnh text-bison. B\u1ea1n c\u00f3 th\u1ec3 t\u1ea1o m\u1ed9t m\u00f4 h\u00ecnh text-bison t\u1eeb xa trong BigQuery b\u1eb1ng m\u1ed9t c\u00e2u l\u1ec7nh SQL nh\u01b0 c\u00e2u l\u1ec7nh d\u01b0\u1edbi \u0111\u00e2y.<\/span><\/p>\n<p>CREATE OR REPLACE MODEL<br \/>\n`bqml_tutorial.text_bison_001` REMOTE<br \/>\nWITH CONNECTION `LOCATION. ConnectionID`<br \/>\nOPTIONS (ENDPOINT =&#8217;text-bison@001&#8242;)<\/p>\n<p><span style=\"font-weight: 400;\">\u0110\u1ec3 suy lu\u1eadn v\u1ec1 m\u00f4 h\u00ecnh, tr\u01b0\u1edbc ti\u00ean Google x\u00e2y d\u1ef1ng m\u1ed9t prompt b\u1eb1ng c\u00e1ch k\u1ebft h\u1ee3p m\u00f4 t\u1ea3 t\u00e1c v\u1ee5 cho m\u00f4 h\u00ecnh\u00a0 v\u00e0 b\u1ea3ng \u0111i\u1ec3m t\u1eeb c\u00e1c b\u1ea3ng m\u00e0 ch\u00ednh Google \u0111\u00e3 t\u1ea1o. Sau \u0111\u00f3, h\u1ecd s\u1eed d\u1ee5ng ch\u1ee9c n\u0103ng <\/span><a href=\"https:\/\/cloud.google.com\/bigquery\/docs\/reference\/standard-sql\/bigqueryml-syntax-generate-text#text-bison\" target=\"_blank\" rel=\"noopener\"><span style=\"font-weight: 400;\">ML.Generate_Text<\/span><\/a><span style=\"font-weight: 400;\"> \u0111\u1ec3 c\u00f3 \u0111\u01b0\u1ee3c \u0111\u1ea7u ra. M\u1eb7c d\u00f9 m\u00f4 h\u00ecnh nh\u1eadn \u0111\u01b0\u1ee3c nhi\u1ec1u ph\u00e2n lo\u1ea1i ch\u00ednh x\u00e1c, nh\u01b0ng n\u00f3 ph\u00e2n lo\u1ea1i m\u1ed9t s\u1ed1 b\u1ea3ng \u0111i\u1ec3m m\u1ed9t c\u00e1ch sai l\u1ea7m. \u1ede \u0111\u00e2y, m\u1ed9t ph\u1ea3n h\u1ed3i m\u1eabu trong \u0111\u00f3 ph\u00e2n lo\u1ea1i kh\u00f4ng ch\u00ednh x\u00e1c.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Prompt<\/span><\/p>\n<p><span style=\"font-weight: 400;\">\u200b<\/span><span style=\"font-weight: 400;\">Please assign a label for the given medical transcript from among these labels [Allergy \/ Immunology, Autopsy, Bariatrics, Cardiovascular \/ Pulmonary, Chiropractic, Consult &#8211; History and Phy., Cosmetic \/ Plastic Surgery, Dentistry, Dermatology, Diets and Nutritions, Discharge Summary, ENT &#8211; Otolaryngology, Emergency Room Reports, Endocrinology, Gastroenterology, General Medicine, Hematology &#8211; Oncology, Hospice &#8211; Palliative Care, IME-QME-Work Comp etc., Lab Medicine &#8211; Pathology, Letters, Nephrology, Neurology, Neurosurgery, Obstetrics \/ Gynecology, Office Notes, Ophthalmology, Orthopedic, Pain Management, Pediatrics &#8211; Neonatal, Physical Medicine &#8211; Rehab, Podiatry, Psychiatry \/ Psychology, Radiology, Rheumatology, SOAP \/ Chart \/ Progress Notes, Sleep Medicine, Speech &#8211; Language, Surgery, Urology]. TRANSCRIPT:<\/span><\/p>\n<p><span style=\"font-weight: 400;\">INDICATIONS FOR PROCEDURE:, The patient has presented with atypical type right arm discomfort and neck discomfort. She had noninvasive vascular imaging demonstrating suspected right subclavian stenosis. Of note, there was bidirectional flow in the right vertebral artery, as well as 250 cm per second velocities in the right subclavian. Duplex ultrasound showed at least a 50% stenosis.,APPROACH:, Right common femoral artery.,ANESTHESIA:, IV sedation with cardiac catheterization protocol. Local infiltration with 1% Xylocaine.,COMPLICATIONS:, None.,ESTIMATED BLOOD LOSS:, Less than 10 ml.,ESTIMATED CONTRAST:, Less than 250 ml.,PROCEDURE PERFORMED:, Right brachiocephalic angiography, right subclavian angiography, selective catheterization of the right subclavian, selective aortic arch angiogram, right iliofemoral angiogram, 6 French Angio-Seal placement.,DESCRIPTION OF PROCEDURE:, The patient was brought to the cardiac catheterization lab in the usual fasting state. She was laid supine on the cardiac catheterization table, and the right groin was prepped and draped in the usual sterile fashion. 1% Xylocaine was infiltrated into the right femoral vessels. Next, a #6 French sheath was introduced into the right femoral artery via the modified Seldinger technique.,AORTIC ARCH ANGIOGRAM:, Next, a pigtail catheter was advanced to the aortic arch. Aortic arch angiogram was then performed with injection of 45 ml of contrast, rate of 20 ml per second, maximum pressure 750 PSI in the 4 degree LAO view.,SELECTIVE SUBCLAVIAN ANGIOGRAPHY:, Next, the right subclavian was selectively cannulated. It was injected in the standard AP, as well as the RAO view. Next pull back pressures were measured across the right subclavian stenosis. No significant gradient was measured.,ANGIOGRAPHIC DETAILS:, The right brachiocephalic artery was patent. The proximal portion of the right carotid was patent. The proximal portion of the right subclavian prior to the origin of the vertebral and the internal mammary showed 50% stenosis.,IMPRESSION:,1. Moderate grade stenosis in the right subclavian artery.,2. Patent proximal edge of the right carotid.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">\u200b<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Response<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Radiology<\/span><\/p>\n<p>&nbsp;<\/p>\n<p><span style=\"font-weight: 400;\">Trong tr\u01b0\u1eddng h\u1ee3p tr\u00ean, ph\u00e2n lo\u1ea1i ch\u00ednh x\u00e1c n\u00ean l\u00e0 \u2018tim m\u1ea1ch\/ ph\u1ed5i.<\/span><\/p>\n<p><b>\u0110\u00e1nh gi\u00e1 d\u1ef1a tr\u00ean c\u00e1c ch\u1ec9 s\u1ed1 cho m\u00f4 h\u00ecnh c\u01a1 s\u1edf<\/b><span style=\"font-weight: 400;\"> \u0111\u1ec3 th\u1ef1c hi\u1ec7n m\u1ed9t \u0111\u00e1nh gi\u00e1 m\u1ea1nh m\u1ebd h\u01a1n v\u1ec1 hi\u1ec7u su\u1ea5t c\u1ee7a m\u00f4 h\u00ecnh, b\u1ea1n c\u00f3 th\u1ec3 s\u1eed d\u1ee5ng h\u00e0m <\/span><a href=\"https:\/\/cloud.google.com\/bigquery\/docs\/reference\/standard-sql\/bigqueryml-syntax-evaluate\" target=\"_blank\" rel=\"noopener\"><span style=\"font-weight: 400;\">ML.EVALUATE<\/span><\/a><span style=\"font-weight: 400;\"> c\u1ee7a BigQuery \u0111\u1ec3 t\u00ednh to\u00e1n c\u00e1c ch\u1ec9 s\u1ed1 v\u1ec1 c\u00e1ch c\u00e1c ph\u1ea3n \u1ee9ng c\u1ee7a m\u00f4 h\u00ecnh so s\u00e1nh v\u1edbi c\u00e1c ph\u1ea3n \u1ee9ng l\u00fd t\u01b0\u1edfng t\u1eeb m\u1ed9t t\u1eadp d\u1eef li\u1ec7u th\u1eed nghi\u1ec7m\/\u0111\u00e1nh gi\u00e1. B\u1ea1n c\u00f3 th\u1ec3 l\u00e0m nh\u01b0 sau:<\/span><\/p>\n<p>&#8212; Evaluate base model<\/p>\n<p>SELECT<br \/>\n*<br \/>\nFROM<br \/>\nml.evaluate(MODEL bqml_tutorial.text_bison_001,<br \/>\n(<br \/>\nSELECT<br \/>\nCONCAT(&#8220;Please assign a label for the given medical transcript from among these labels [Allergy \/ Immunology, Autopsy, Bariatrics, Cardiovascular \/ Pulmonary, Chiropractic, Consult &#8211; History and Phy., Cosmetic \/ Plastic Surgery, Dentistry, Dermatology, Diets and Nutritions, Discharge Summary, ENT &#8211; Otolaryngology, Emergency Room Reports, Endocrinology, Gastroenterology, General Medicine, Hematology &#8211; Oncology, Hospice &#8211; Palliative Care, IME-QME-Work Comp etc., Lab Medicine &#8211; Pathology, Letters, Nephrology, Neurology, Neurosurgery, Obstetrics \/ Gynecology, Office Notes, Ophthalmology, Orthopedic, Pain Management, Pediatrics &#8211; Neonatal, Physical Medicine &#8211; Rehab, Podiatry, Psychiatry \/ Psychology, Radiology, Rheumatology, SOAP \/ Chart \/ Progress Notes, Sleep Medicine, Speech &#8211; Language, Surgery, Urology]. &#8220;, input_text) AS input_text,<br \/>\noutput_text<br \/>\nFROM<br \/>\n`bqml_tutorial.medical_transcript_eval` ),<br \/>\nSTRUCT(&#8220;classification&#8221; AS task_type))<\/p>\n<p><span style=\"font-weight: 400;\">Trong m\u00e3 tr\u00ean, Google \u0111\u00e3 cung c\u1ea5p m\u1ed9t b\u1ea3ng \u0111\u00e1nh gi\u00e1 l\u00e0m \u0111\u1ea7u v\u00e0o v\u00e0 ch\u1ecdn \u2018Ph\u00e2n lo\u1ea1i\u2018 l\u00e0 lo\u1ea1i t\u00e1c v\u1ee5 m\u00e0 h\u1ecd \u0111\u00e1nh gi\u00e1 m\u00f4 h\u00ecnh. Google \u0111\u1ec3 l\u1ea1i c\u00e1c <\/span><a href=\"https:\/\/cloud.google.com\/bigquery\/docs\/generate-text-tuning#generate_text\" target=\"_blank\" rel=\"noopener\"><span style=\"font-weight: 400;\">th\u00f4ng s\u1ed1 suy lu\u1eadn<\/span><\/a><span style=\"font-weight: 400;\"> kh\u00e1c theo m\u1eb7c \u0111\u1ecbnh c\u1ee7a ch\u00fang nh\u01b0ng ch\u00fang c\u00f3 th\u1ec3 \u0111\u01b0\u1ee3c s\u1eeda \u0111\u1ed5i \u0111\u1ec3 \u0111\u00e1nh gi\u00e1.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">C\u00e1c s\u1ed1 li\u1ec7u \u0111\u00e1nh gi\u00e1 \u0111\u01b0\u1ee3c tr\u1ea3 v\u1ec1 \u0111\u01b0\u1ee3c t\u00ednh to\u00e1n cho m\u1ed7i l\u1edbp (nh\u00e3n). K\u1ebft qu\u1ea3 tr\u00f4ng gi\u1ed1ng nh\u01b0 sau:<\/span><\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-18428 size-full\" src=\"https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry-1.jpg\" alt=\"\" width=\"511\" height=\"585\" srcset=\"https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry-1.jpg 511w, https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry-1-10x12.jpg 10w\" sizes=\"auto, (max-width: 511px) 100vw, 511px\" \/><\/p>\n<p><span style=\"font-weight: 400;\">T\u1eadp trung v\u00e0o <\/span><a href=\"https:\/\/en.wikipedia.org\/wiki\/F-score\" target=\"_blank\" rel=\"noopener\"><span style=\"font-weight: 400;\">\u0111i\u1ec3m F1<\/span><\/a><span style=\"font-weight: 400;\"> (gi\u00e1 tr\u1ecb trung b\u00ecnh h\u00f2a c\u1ee7a precision v\u00e0 recall), b\u1ea1n c\u00f3 th\u1ec3 th\u1ea5y r\u1eb1ng hi\u1ec7u su\u1ea5t c\u1ee7a m\u00f4 h\u00ecnh thay \u0111\u1ed5i gi\u1eefa c\u00e1c l\u1edbp. V\u00ed d\u1ee5, m\u00f4 h\u00ecnh c\u01a1 s\u1edf ho\u1ea1t \u0111\u1ed9ng t\u1ed1t cho c\u00e1c l\u1edbp \u2018Autopsy\u2019, \u2018Diets and Nutritions\u2019, v\u00e0 \u2018Dentistry\u2019, nh\u01b0ng ho\u1ea1t \u0111\u1ed9ng k\u00e9m cho c\u00e1c l\u1edbp \u2018Consult &#8211; History and Phy.\u2019, \u2018Chiropractic\u2019, v\u00e0 \u2018Cardiovascular \/ Pulmonary\u2019.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">B\u00e2y gi\u1edd h\u00e3y tinh ch\u1ec9nh m\u00f4 h\u00ecnh c\u1ee7a ch\u00fang t\u00f4i v\u00e0 xem li\u1ec7u ch\u00fang ta c\u00f3 th\u1ec3 c\u1ea3i thi\u1ec7n hi\u1ec7u su\u1ea5t c\u01a1 s\u1edf n\u00e0y kh\u00f4ng.<\/span><\/p>\n<h2><span class=\"ez-toc-section\" id=\"Tao_mot_mo_hinh_duoc_tinh_chinh\"><\/span><b>T\u1ea1o m\u1ed9t m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh<\/b><span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p><span style=\"font-weight: 400;\">Vi\u1ec7c t\u1ea1o m\u1ed9t m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh trong BigQuery l\u00e0 \u0111\u01a1n gi\u1ea3n. B\u1ea1n c\u00f3 th\u1ec3 th\u1ef1c hi\u1ec7n tinh ch\u1ec9nh b\u1eb1ng c\u00e1ch ch\u1ec9 \u0111\u1ecbnh d\u1eef li\u1ec7u hu\u1ea5n luy\u1ec7n v\u1edbi c\u00e1c c\u1ed9t &#8216;prompt&#8217; v\u00e0 &#8216;label&#8217; trong c\u00e2u l\u1ec7nh T\u1ea1o M\u00f4 h\u00ecnh. Ch\u00fang ta s\u1eed d\u1ee5ng c\u00f9ng m\u1ed9t prompt \u0111\u1ec3 tinh ch\u1ec9nh m\u00e0 ch\u00fang ta \u0111\u00e3 s\u1eed d\u1ee5ng trong \u0111\u00e1nh gi\u00e1 tr\u01b0\u1edbc \u0111\u00f3. T\u1ea1o m\u1ed9t m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh nh\u01b0 sau:<\/span><\/p>\n<p>&#8212; Fine tune a textbison model<\/p>\n<p>CREATE OR REPLACE MODEL<br \/>\n`bqml_tutorial.text_bison_001_medical_transcript_finetuned` REMOTE<br \/>\nWITH CONNECTION `LOCATION. ConnectionID`<br \/>\nOPTIONS (endpoint=&#8221;text-bison@001&#8243;,<br \/>\nmax_iterations=300,<br \/>\ndata_split_method=&#8221;no_split&#8221;) AS<br \/>\nSELECT<br \/>\nCONCAT(&#8220;Please assign a label for the given medical transcript from among these labels [Allergy \/ Immunology, Autopsy, Bariatrics, Cardiovascular \/ Pulmonary, Chiropractic, Consult &#8211; History and Phy., Cosmetic \/ Plastic Surgery, Dentistry, Dermatology, Diets and Nutritions, Discharge Summary, ENT &#8211; Otolaryngology, Emergency Room Reports, Endocrinology, Gastroenterology, General Medicine, Hematology &#8211; Oncology, Hospice &#8211; Palliative Care, IME-QME-Work Comp etc., Lab Medicine &#8211; Pathology, Letters, Nephrology, Neurology, Neurosurgery, Obstetrics \/ Gynecology, Office Notes, Ophthalmology, Orthopedic, Pain Management, Pediatrics &#8211; Neonatal, Physical Medicine &#8211; Rehab, Podiatry, Psychiatry \/ Psychology, Radiology, Rheumatology, SOAP \/ Chart \/ Progress Notes, Sleep Medicine, Speech &#8211; Language, Surgery, Urology]. &#8220;, input_text) AS prompt,<br \/>\noutput_text AS label<br \/>\nFROM<br \/>\n`bqml_tutorial.medical_transcript_train`<\/p>\n<p><span style=\"font-weight: 400;\">C\u00c1C K\u1ebeT N\u1ed0I b\u1ea1n s\u1eed d\u1ee5ng \u0111\u1ec3 t\u1ea1o m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh ph\u1ea3i c\u00f3 (a) Storage Object User v\u00e0 (b) Vertex AI Service Agent roles. Ngo\u00e0i ra, t\u00e0i kho\u1ea3n d\u1ecbch v\u1ee5 m\u1eb7c \u0111\u1ecbnh c\u1ee7a Compute Engine (GCE) c\u1ee7a b\u1ea1n ph\u1ea3i c\u00f3 quy\u1ec1n <\/span><a href=\"https:\/\/cloud.google.com\/bigquery\/docs\/generate-text-tuning#gce-service-account-access\" target=\"_blank\" rel=\"noopener\"><span style=\"font-weight: 400;\">ch\u1ec9nh s\u1eeda <\/span><\/a><span style=\"font-weight: 400;\">\u00a0\u0111\u1ed1i v\u1edbi d\u1ef1 \u00e1n. Tham kh\u1ea3o <\/span><a href=\"https:\/\/cloud.google.com\/bigquery\/docs\/generate-text-tutorial#create_a_connection\" target=\"_blank\" rel=\"noopener\"><span style=\"font-weight: 400;\">t\u00e0i li\u1ec7u<\/span><\/a><span style=\"font-weight: 400;\"> \u0111\u1ec3 bi\u1ebft h\u01b0\u1edbng d\u1eabn v\u1ec1 c\u00e1ch l\u00e0m vi\u1ec7c v\u1edbi c\u00e1c k\u1ebft n\u1ed1i BigQuery.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">BigQuery th\u1ef1c hi\u1ec7n vi\u1ec7c tinh ch\u1ec9nh m\u00f4 h\u00ecnh b\u1eb1ng c\u00e1ch s\u1eed d\u1ee5ng m\u1ed9t k\u1ef9 thu\u1eadt \u0111\u01b0\u1ee3c bi\u1ebft \u0111\u1ebfn l\u00e0 Low-Rank Adaptation (LoRA). Vi\u1ec7c \u0111i\u1ec1u ch\u1ec9nh LoRA l\u00e0 m\u1ed9t ph\u01b0\u01a1ng ph\u00e1p \u0111i\u1ec1u ch\u1ec9nh hi\u1ec7u qu\u1ea3 tham s\u1ed1 (PET) m\u00e0 \u0111\u00f3ng b\u0103ng tr\u1ecdng s\u1ed1 m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c \u0111\u00e0o t\u1ea1o tr\u01b0\u1edbc v\u00e0 ghi \u0111\u00e8 ma tr\u1eadn ph\u00e2n t\u00e1ch h\u1ea1ng c\u00f3 th\u1ec3 \u0111\u00e0o t\u1ea1o v\u00e0o m\u1ed7i l\u1edbp c\u1ee7a ki\u1ebfn tr\u00fac Transformer \u0111\u1ec3 <\/span><a href=\"https:\/\/arxiv.org\/pdf\/2106.09685.pdf\" target=\"_blank\" rel=\"noopener\"><span style=\"font-weight: 400;\">gi\u1ea3m s\u1ed1 l\u01b0\u1ee3ng tham s\u1ed1 c\u00f3 th\u1ec3 \u0111\u00e0o t\u1ea1o<\/span><\/a><span style=\"font-weight: 400;\">. Vi\u1ec7c tinh ch\u1ec9nh m\u00f4 h\u00ecnh x\u1ea3y ra tr\u00ean m\u1ed9t m\u00e1y t\u00ednh Vertex AI v\u00e0 b\u1ea1n c\u00f3 th\u1ec3 ch\u1ecdn GPU ho\u1eb7c TPU l\u00e0m b\u1ed9 gia t\u1ed1c. B\u1ea1n s\u1ebd \u0111\u01b0\u1ee3c t\u00ednh ph\u00ed b\u1edfi BigQuery cho d\u1eef li\u1ec7u \u0111\u01b0\u1ee3c qu\u00e9t ho\u1eb7c khe c\u1eafm s\u1eed d\u1ee5ng, c\u0169ng nh\u01b0 b\u1edfi Vertex AI cho c\u00e1c t\u00e0i nguy\u00ean Vertex AI \u0111\u01b0\u1ee3c ti\u00eau th\u1ee5. C\u00f4ng vi\u1ec7c tinh ch\u1ec9nh t\u1ea1o ra m\u1ed9t \u0111i\u1ec3m cu\u1ed1i m\u00f4 h\u00ecnh m\u1edbi \u0111\u1ea1i di\u1ec7n cho c\u00e1c tr\u1ecdng s\u1ed1 \u0111\u00e3 h\u1ecdc. C\u00e1c kho\u1ea3n ph\u00ed d\u1ef1 \u0111o\u00e1n c\u1ee7a Vertex AI m\u00e0 b\u1ea1n ph\u1ea3i tr\u1ea3 khi truy v\u1ea5n m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh l\u00e0 gi\u1ed1ng nh\u01b0 v\u1edbi m\u00f4 h\u00ecnh c\u01a1 s\u1edf.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">C\u00f4ng vi\u1ec7c tinh ch\u1ec9nh n\u00e0y c\u00f3 th\u1ec3 m\u1ea5t m\u1ed9t v\u00e0i gi\u1edd \u0111\u1ec3 ho\u00e0n th\u00e0nh, thay \u0111\u1ed5i d\u1ef1a tr\u00ean c\u00e1c t\u00f9y ch\u1ecdn hu\u1ea5n luy\u1ec7n nh\u01b0 &#8216;max_iterations&#8217;. Sau khi ho\u00e0n th\u00e0nh, b\u1ea1n c\u00f3 th\u1ec3 t\u00ecm th\u1ea5y chi ti\u1ebft c\u1ee7a m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh c\u1ee7a b\u1ea1n trong giao di\u1ec7n ng\u01b0\u1eddi d\u00f9ng BigQuery, n\u01a1i b\u1ea1n s\u1ebd th\u1ea5y m\u1ed9t \u0111i\u1ec3m cu\u1ed1i t\u1eeb xa kh\u00e1c cho m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh.<\/span><\/p>\n<p><span style=\"font-weight: 400;\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-18427 size-full\" src=\"https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry-2.jpg\" alt=\"\" width=\"601\" height=\"243\" srcset=\"https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry-2.jpg 601w, https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry-2-18x7.jpg 18w\" sizes=\"auto, (max-width: 601px) 100vw, 601px\" \/><\/span><\/p>\n<p><span style=\"font-weight: 400;\">Hi\u1ec7n t\u1ea1i, BigQuery h\u1ed7 tr\u1ee3 vi\u1ec7c tinh ch\u1ec9nh c\u1ee7a c\u00e1c m\u00f4 h\u00ecnh text-bison-001 v\u00e0 text-bison-002.<\/span><\/p>\n<h2><span class=\"ez-toc-section\" id=\"Danh_gia_hieu_xuat_tinh_chinh_model\"><\/span><b>\u0110\u00e1nh gi\u00e1 hi\u1ec7u xu\u1ea5t tinh ch\u1ec9nh model<\/b><span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p><span style=\"font-weight: 400;\">B\u00e2y gi\u1edd b\u1ea1n c\u00f3 th\u1ec3 t\u1ea1o d\u1ef1 \u0111o\u00e1n t\u1eeb vi\u1ec7c tinh ch\u1ec9nh model s\u1eed d\u1ee5ng code nh\u01b0 sau :\u00a0<\/span><\/p>\n<p>SELECT<br \/>\nml_generate_text_llm_result,<br \/>\nlabel,<br \/>\nprompt<br \/>\nFROM<br \/>\nml.generate_text(MODEL bqml_tutorial.text_bison_001_medical_transcript_finetuned,<br \/>\n(<br \/>\nSELECT<br \/>\nCONCAT(&#8220;Please assign a label for the given medical transcript from among these labels [Allergy \/ Immunology, Autopsy, Bariatrics, Cardiovascular \/ Pulmonary, Chiropractic, Consult &#8211; History and Phy., Cosmetic \/ Plastic Surgery, Dentistry, Dermatology, Diets and Nutritions, Discharge Summary, ENT &#8211; Otolaryngology, Emergency Room Reports, Endocrinology, Gastroenterology, General Medicine, Hematology &#8211; Oncology, Hospice &#8211; Palliative Care, IME-QME-Work Comp etc., Lab Medicine &#8211; Pathology, Letters, Nephrology, Neurology, Neurosurgery, Obstetrics \/ Gynecology, Office Notes, Ophthalmology, Orthopedic, Pain Management, Pediatrics &#8211; Neonatal, Physical Medicine &#8211; Rehab, Podiatry, Psychiatry \/ Psychology, Radiology, Rheumatology, SOAP \/ Chart \/ Progress Notes, Sleep Medicine, Speech &#8211; Language, Surgery, Urology]. &#8220;, input_text) AS prompt,<br \/>\noutput_text as label<br \/>\nFROM<br \/>\n`bqml_tutorial.medical_transcript_eval`<br \/>\n),<br \/>\nSTRUCT(TRUE AS flatten_json_output))<\/p>\n<p><span style=\"font-weight: 400;\">H\u00e3y xem x\u00e9t ph\u1ea3n h\u1ed3i \u0111\u1ed1i v\u1edbi prompt m\u1eabu m\u00e0 Google \u0111\u00e3 \u0111\u00e1nh gi\u00e1 tr\u01b0\u1edbc \u0111\u00f3. S\u1eed d\u1ee5ng c\u00f9ng m\u1ed9t prompt, m\u00f4 h\u00ecnh b\u00e2y gi\u1edd ph\u00e2n lo\u1ea1i b\u1ea3n ghi nh\u01b0 \u2018Cardiovascular \/ Pulmonary\u2019 \u2014 ph\u1ea3n h\u1ed3i ch\u00ednh x\u00e1c.<\/span><\/p>\n<h3><span class=\"ez-toc-section\" id=\"Danh_gia_dua_tren_chi_so_cho_mo_hinh_duoc_tinh_chinh\"><\/span><b>\u0110\u00e1nh gi\u00e1 d\u1ef1a tr\u00ean ch\u1ec9 s\u1ed1 cho m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh<\/b><span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p><span style=\"font-weight: 400;\">B\u00e2y gi\u1edd, ch\u00fang ta s\u1ebd t\u00ednh to\u00e1n c\u00e1c ch\u1ec9 s\u1ed1 cho m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh b\u1eb1ng c\u00e1ch s\u1eed d\u1ee5ng c\u00f9ng m\u1ed9t d\u1eef li\u1ec7u \u0111\u00e1nh gi\u00e1 v\u00e0 c\u00f9ng m\u1ed9t prompt m\u00e0 Google \u0111\u00e3 s\u1eed d\u1ee5ng tr\u01b0\u1edbc \u0111\u00f3 \u0111\u1ec3 \u0111\u00e1nh gi\u00e1 m\u00f4 h\u00ecnh c\u01a1 s\u1edf.<\/span><\/p>\n<p>&#8212; Evaluate fine tuned model<\/p>\n<p>SELECT<br \/>\n*<br \/>\nFROM<br \/>\nml.evaluate(MODEL bqml_tutorial.text_bison_001_medical_transcript_finetuned,<br \/>\n(<br \/>\nSELECT<br \/>\nCONCAT(&#8220;Please assign a label for the given medical transcript from among these labels [Allergy \/ Immunology, Autopsy, Bariatrics, Cardiovascular \/ Pulmonary, Chiropractic, Consult &#8211; History and Phy., Cosmetic \/ Plastic Surgery, Dentistry, Dermatology, Diets and Nutritions, Discharge Summary, ENT &#8211; Otolaryngology, Emergency Room Reports, Endocrinology, Gastroenterology, General Medicine, Hematology &#8211; Oncology, Hospice &#8211; Palliative Care, IME-QME-Work Comp etc., Lab Medicine &#8211; Pathology, Letters, Nephrology, Neurology, Neurosurgery, Obstetrics \/ Gynecology, Office Notes, Ophthalmology, Orthopedic, Pain Management, Pediatrics &#8211; Neonatal, Physical Medicine &#8211; Rehab, Podiatry, Psychiatry \/ Psychology, Radiology, Rheumatology, SOAP \/ Chart \/ Progress Notes, Sleep Medicine, Speech &#8211; Language, Surgery, Urology]. &#8220;, input_text) AS prompt,<br \/>\noutput_text as label<br \/>\nFROM<br \/>\n`bqml_tutorial.medical_transcript_eval`), STRUCT(&#8220;classification&#8221; AS task_type))<\/p>\n<p><span style=\"font-weight: 400;\">C\u00e1c ch\u1ec9 s\u1ed1 t\u1eeb m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh \u0111\u01b0\u1ee3c hi\u1ec3n th\u1ecb d\u01b0\u1edbi \u0111\u00e2y. M\u1eb7c d\u00f9 t\u1eadp d\u1eef li\u1ec7u tinh ch\u1ec9nh (hu\u1ea5n luy\u1ec7n) m\u00e0 Google s\u1eed d\u1ee5ng cho blog n\u00e0y ch\u1ec9 ch\u1ee9a 519 v\u00ed d\u1ee5, ch\u00fang ta \u0111\u00e3 th\u1ea5y m\u1ed9t c\u1ea3i thi\u1ec7n \u0111\u00e1ng k\u1ec3 trong hi\u1ec7u su\u1ea5t. \u0110i\u1ec3m F1 tr\u00ean c\u00e1c nh\u00e3n, n\u01a1i m\u00e0 m\u00f4 h\u00ecnh \u0111\u00e3 ho\u1ea1t \u0111\u1ed9ng k\u00e9m tr\u01b0\u1edbc \u0111\u00e2y, \u0111\u00e3 \u0111\u01b0\u1ee3c c\u1ea3i thi\u1ec7n, v\u1edbi \u0111i\u1ec3m F1 &#8220;macro&#8221; (trung b\u00ecnh \u0111\u01a1n gi\u1ea3n c\u1ee7a \u0111i\u1ec3m F1 tr\u00ean t\u1ea5t c\u1ea3 c\u00e1c nh\u00e3n) t\u0103ng t\u1eeb 0,54 l\u00ean 0,66.<\/span><\/p>\n<h2><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-18426 size-full\" src=\"https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry-3.jpg\" alt=\"\" width=\"521\" height=\"602\" srcset=\"https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry-3.jpg 521w, https:\/\/gcloudvn.com\/wp-content\/uploads\/2024\/04\/llm-bigquerry-3-10x12.jpg 10w\" sizes=\"auto, (max-width: 521px) 100vw, 521px\" \/><br \/>\n<b>S\u1eb5n s\u00e0ng cho suy lu\u1eadn\u00a0<\/b><\/h2>\n<p><span style=\"font-weight: 400;\">M\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh b\u00e2y gi\u1edd c\u00f3 th\u1ec3 \u0111\u01b0\u1ee3c s\u1eed d\u1ee5ng cho vi\u1ec7c suy lu\u1eadn b\u1eb1ng c\u00e1ch s\u1eed d\u1ee5ng h\u00e0m ML.GENERATE_TEXT, m\u00e0 Google \u0111\u00e3 s\u1eed d\u1ee5ng trong c\u00e1c b\u01b0\u1edbc tr\u01b0\u1edbc \u0111\u1ec3 c\u00f3 \u0111\u01b0\u1ee3c c\u00e1c ph\u1ea3n h\u1ed3i m\u1eabu. B\u1ea1n kh\u00f4ng c\u1ea7n qu\u1ea3n l\u00fd b\u1ea5t k\u1ef3 c\u01a1 s\u1edf h\u1ea1 t\u1ea7ng b\u1ed5 sung n\u00e0o cho m\u00f4 h\u00ecnh \u0111\u01b0\u1ee3c tinh ch\u1ec9nh c\u1ee7a b\u1ea1n v\u00e0 b\u1ea1n \u0111\u01b0\u1ee3c t\u00ednh gi\u00e1 suy lu\u1eadn gi\u1ed1ng nh\u01b0 b\u1ea1n \u0111\u00e3 ph\u1ea3i ch\u1ecbu cho m\u00f4 h\u00ecnh c\u01a1 s\u1edf.<\/span><\/p>\n\n\t\t<\/div>\n\t<\/div>\n<div class=\"templatera_shortcode\"><div class=\"vc_row wpb_row vc_row-fluid\"><div class=\"wpb_column vc_column_container vc_col-sm-12\"><div class=\"vc_column-inner\"><div class=\"wpb_wrapper\"><div class=\"vc_message_box vc_message_box-standard vc_message_box-rounded vc_color-blue\" ><div class=\"vc_message_box-icon\"><i class=\"vc-mono vc-mono-technorati\"><\/i><\/div><p><a href=\"https:\/\/gcloudvn.com\/main-logo-1\/\"><img loading=\"lazy\" decoding=\"async\" class=\"alignleft wp-image-664\" src=\"https:\/\/gcloudvn.com\/wp-content\/uploads\/2021\/06\/main-logo-1.png\" alt=\"\" width=\"221\" height=\"72\" srcset=\"https:\/\/gcloudvn.com\/wp-content\/uploads\/2021\/06\/main-logo-1.png 214w, https:\/\/gcloudvn.com\/wp-content\/uploads\/2021\/06\/main-logo-1-18x6.png 18w, https:\/\/gcloudvn.com\/wp-content\/uploads\/2021\/06\/main-logo-1-183x60.png 183w\" sizes=\"auto, (max-width: 221px) 100vw, 221px\" \/><\/a>L\u00e0 \u0111\u1ed1i t\u00e1c c\u1ea5p cao c\u1ee7a Google t\u1ea1i Vi\u1ec7t Nam, Gimasys c\u00f3 h\u01a1n 10+ n\u0103m kinh nghi\u1ec7m, t\u01b0 v\u1ea5n tri\u1ec3n khai chuy\u1ec3n \u0111\u1ed1i s\u1ed1 cho 2000+ doanh nghi\u1ec7p t\u1eadp \u0111o\u00e0n trong n\u01b0\u1edbc. M\u1ed9t s\u1ed1 kh\u00e1ch h\u00e0ng ti\u00eau bi\u1ec3u Jetstar, \u0110i\u1ec1n Qu\u00e2n Media, Heineken, Jollibee, Vietnam Airline, HSC, SSI...<\/p>\n<p>Gimasys hi\u1ec7n \u0111ang l\u00e0 \u0111\u1ed1i t\u00e1c chi\u1ebfn l\u01b0\u1ee3c c\u1ee7a h\u00e0ng lo\u1ea1t h\u00e3ng c\u00f4ng ngh\u1ec7 l\u1edbn tr\u00ean th\u1ebf gi\u1edbi nh\u01b0 Salesforce, Oracle Netsuite, Tableau, Mulesoft<\/p>\n<p>Li\u00ean h\u1ec7 Gimasys - Google Cloud Premier Partner \u0111\u1ec3 \u0111\u01b0\u1ee3c t\u01b0 v\u1ea5n c\u00e1c gi\u1ea3i ph\u00e1p chi\u1ebfn l\u01b0\u1ee3c ph\u00f9 h\u1ee3p nhu c\u1ea7u ri\u00eang c\u1ee7a doanh nghi\u1ec7p:<\/p>\n<ul>\n<li>Email: gcp@gimasys.com<\/li>\n<li>Hotline: 0974 417 099<\/li>\n<\/ul>\n<\/div><\/div><\/div><\/div><\/div><\/div><\/div><\/div><\/div><\/div>\n<\/section>","protected":false},"excerpt":{"rendered":"BigQuery cho ph\u00e9p b\u1ea1n ph\u00e2n t\u00edch d\u1eef li\u1ec7u c\u1ee7a m\u00ecnh b\u1eb1ng c\u00e1ch s\u1eed d\u1ee5ng m\u1ed9t lo\u1ea1t c\u00e1c m\u00f4 h\u00ecnh ng\u00f4n ng\u1eef l\u1edbn (LLMs) \u0111\u01b0\u1ee3c l\u01b0u tr\u1eef trong Vertex AI bao g\u1ed3m Gemini 1.0 Pro, Gemini 1.0 Pro Vision v\u00e0 text-bison.&hellip;","protected":false},"author":2,"featured_media":18425,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"inline_featured_image":false,"footnotes":""},"categories":[1],"tags":[],"class_list":["post-18435","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-kienthuc","entry","has-media"],"_links":{"self":[{"href":"https:\/\/gcloudvn.com\/en\/wp-json\/wp\/v2\/posts\/18435","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/gcloudvn.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/gcloudvn.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/gcloudvn.com\/en\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/gcloudvn.com\/en\/wp-json\/wp\/v2\/comments?post=18435"}],"version-history":[{"count":0,"href":"https:\/\/gcloudvn.com\/en\/wp-json\/wp\/v2\/posts\/18435\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/gcloudvn.com\/en\/wp-json\/wp\/v2\/media\/18425"}],"wp:attachment":[{"href":"https:\/\/gcloudvn.com\/en\/wp-json\/wp\/v2\/media?parent=18435"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/gcloudvn.com\/en\/wp-json\/wp\/v2\/categories?post=18435"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/gcloudvn.com\/en\/wp-json\/wp\/v2\/tags?post=18435"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}