{"id":5517,"date":"2025-02-19T10:09:18","date_gmt":"2025-02-19T10:09:18","guid":{"rendered":"https:\/\/www.myshirtai.com\/?p=5517"},"modified":"2025-02-19T10:23:53","modified_gmt":"2025-02-19T10:23:53","slug":"%e6%9c%ac%e5%9c%b0deepseek%e6%a8%a1%e5%9e%8b%e9%9c%80%e8%a6%81%e4%bb%80%e4%b9%88%e9%85%8d%e7%bd%ae%e4%bb%a5%e5%8f%8a%e5%90%84%e9%85%8d%e7%bd%ae%e8%b7%91%e5%88%86","status":"publish","type":"post","link":"https:\/\/www.myshirtai.com\/pt\/archives\/5517","title":{"rendered":"Que configura\u00e7\u00f5es s\u00e3o necess\u00e1rias para o modelo DeepSeek local e as pontua\u00e7\u00f5es de tempo de execu\u00e7\u00e3o para cada configura\u00e7\u00e3o"},"content":{"rendered":"<h4>I. Conclus\u00f5es do estudo<\/h4>\n<h5>1. conclus\u00f5es gerais<\/h5>\n<p>Os resultados deste estudo mostram que a execu\u00e7\u00e3o da vers\u00e3o b\u00e1sica do modelo DeepSeek nas condi\u00e7\u00f5es de maior capacidade de computa\u00e7\u00e3o que podem atualmente ser encontradas localmente ainda enfrenta desafios significativos. Especificamente, o custo de constru\u00e7\u00e3o \u00e9 demasiado elevado e ainda n\u00e3o \u00e9 suficiente para suportar cen\u00e1rios gerais, tais como perguntas e respostas cont\u00ednuas e apoio ao desenvolvimento em termos de desempenho e qualidade.<\/p>\n<p>Se algu\u00e9m desejar treinar um modelo especializado com base na vers\u00e3o de base do modelo DeepSeek para aplica\u00e7\u00e3o num produto, \u00e9 necess\u00e1rio considerar cuidadosamente os requisitos t\u00e9cnicos do cen\u00e1rio de aplica\u00e7\u00e3o em termos de simultaneidade, pontualidade, etc. A rela\u00e7\u00e3o entre o tamanho do modelo de base e a aritm\u00e9tica alvo do produto deve ser razoavelmente avaliada de modo a alcan\u00e7ar um equil\u00edbrio entre o custo e a efic\u00e1cia do produto.<\/p>\n<p>Embora existam muitas limita\u00e7\u00f5es no funcionamento do modelo DeepSeek no atual ambiente de hardware local, isso n\u00e3o significa que esteja completamente inexplorado. Se sob a premissa de aumentar adequadamente o custo do hardware, como aumentar a capacidade da mem\u00f3ria de v\u00eddeo e adotar uma arquitetura de hardware mais eficiente, etc., e ao mesmo tempo, meios t\u00e9cnicos como o treino de destila\u00e7\u00e3o com base em modelos mais pequenos como o 7B podem ser refor\u00e7ados para melhorar a qualidade do modelo e satisfazer melhor as necessidades das aplica\u00e7\u00f5es locais. Al\u00e9m disso, tamb\u00e9m \u00e9 poss\u00edvel explorar profundamente a forma de otimizar o algoritmo do modelo e a depura\u00e7\u00e3o de par\u00e2metros para melhorar ainda mais o desempenho do modelo nas condi\u00e7\u00f5es de hardware existentes.<\/p>\n<p><img fetchpriority=\"high\" decoding=\"async\" class=\"aligncenter\" src=\"https:\/\/ollama.com\/assets\/library\/deepseek-r1\/e44d096e-fa46-4cae-b2f2-53991e8c8da0\" alt=\"deepseek\" width=\"1060\" height=\"628\" \/><\/p>\n<h5>2. desempenho de diferentes modelos locais<\/h5>\n<p>Conseguimos suportar at\u00e9 70b de execu\u00e7\u00f5es de modelos do DeepSeek R1 com base nos requisitos m\u00ednimos de configura\u00e7\u00e3o para a implementa\u00e7\u00e3o local dos modelos do s\u00edtio Web do DeepSeek, combinados com o melhor hardware que t\u00ednhamos dispon\u00edvel (ou seja, 2 NVIDIA A100 80G de mem\u00f3ria gr\u00e1fica), e n\u00e3o conseguimos executar o modelo completo de 671b.<\/p>\n<p>Tent\u00e1mos instalar um total de 6 modelos de 70b e inferiores e todos eles funcionaram corretamente. Os modelos de 1,5b n\u00e3o foram eficazes e base\u00e1mos os nossos testes comparativos e an\u00e1lises principalmente nos modelos de 70b e 7b.<\/p>\n<p>Al\u00e9m disso, o primeiro teste realizado com um \u00fanico cart\u00e3o revelou que a velocidade de resposta do modelo 70b \u00e9 demasiado lenta, o teste com dois cart\u00f5es apenas para as diferen\u00e7as de desempenho te\u00f3rico de um \u00fanico cart\u00e3o duplo (o mesmo modelo de impacto aritm\u00e9tico diferente na velocidade de desempenho do racioc\u00ednio, teoricamente n\u00e3o afecta a qualidade, a verifica\u00e7\u00e3o simples tamb\u00e9m est\u00e1 de acordo com o cen\u00e1rio te\u00f3rico), por isso, no ambiente experimental de dois cart\u00f5es, apenas utilizamos o modelo 7b para uma vasta gama de valida\u00e7\u00e3o.<\/p>\n<p><strong>7b<\/strong><strong>Modela\u00e7\u00e3o do desempenho:<\/strong>No teste de 5 pessoas com carga total, o modelo 7b respondeu relativamente r\u00e1pido nas primeiras perguntas e respostas (quase 35 segundos para o cart\u00e3o duplo e quase 70 segundos para o cart\u00e3o simples). A estrutura e a qualidade do conte\u00fado da resposta tiveram um desempenho moderadamente bom, mas depois de fazer algumas perguntas inferenciais complexas ou perguntas de seguimento cont\u00ednuas, devido ao crescimento do contexto, o modelo 7b come\u00e7ou a apresentar respostas incoerentes, inventadas e mal concebidas, embora a velocidade de resposta fosse est\u00e1vel.<\/p>\n<p><strong>70b<\/strong><strong>Modela\u00e7\u00e3o do desempenho:<\/strong>Num teste de carga total com 5 pessoas, o modelo 70b foi muito lento a responder \u00e0 primeira resposta \u00e0 mesma pergunta (mais de 7 minutos para o cart\u00e3o \u00fanico, n\u00e3o testado em pormenor para o cart\u00e3o duplo apenas para valida\u00e7\u00e3o simples). O conte\u00fado da resposta era um pouco melhor do que o do modelo 7b em termos de estrutura, disposi\u00e7\u00e3o e qualidade, mas n\u00e3o estava muito \u00e0 frente das respostas do modelo 7b e, \u00e0 medida que o contexto aumentava (mais longo do que o do modelo 7b), o modelo 70b tamb\u00e9m apresentava os mesmos fen\u00f3menos de m\u00e1 qualidade de resposta, l\u00f3gica confusa e inven\u00e7\u00f5es. Em particular, o tempo de resposta do modelo 70b \u00e9 demasiado longo para o hardware dispon\u00edvel, o que resulta numa m\u00e1 experi\u00eancia do utilizador e afecta seriamente a sua pontua\u00e7\u00e3o de qualidade.<\/p>\n<p>Por \u00faltimo, atrav\u00e9s dos dados de classifica\u00e7\u00e3o dos utilizadores, tanto o modelo 7b como o 70b falharam em termos de qualidade do conte\u00fado da resposta, tendo o modelo 7b um n\u00edvel ligeiramente superior de satisfa\u00e7\u00e3o dos utilizadores devido \u00e0 sua resposta relativamente r\u00e1pida.<\/p>\n<h5>3) Compara\u00e7\u00e3o entre o modelo local 70b e o modelo oficial baseado na Web<\/h5>\n<p>As respostas do modelo 70b s\u00e3o de qualidade m\u00e9dia.<\/p>\n<p>Relativamente \u00e0 qualidade das respostas ao modelo 70b, organiz\u00e1mos v\u00e1rios testes. As mesmas perguntas foram feitas ao modelo DeepSeek-R1:70b implantado localmente e ao s\u00edtio Web oficial do DeepSeek online (ou seja, o modelo DeepSeek-R1 completo).<\/p>\n<p>Em primeiro lugar, existe uma diferen\u00e7a na velocidade de resposta. No modelo local 70b, a velocidade de resposta \u00e9 de cerca de 70 segundos (teste individual), ao passo que na vers\u00e3o oficial da Web a velocidade de resposta \u00e9 de cerca de 30 segundos (teste individual).<\/p>\n<p>Em segundo lugar, existe uma diferen\u00e7a na qualidade do conte\u00fado das respostas entre os dois modelos. O modelo 70b d\u00e1 ocasionalmente respostas simples a perguntas de conhecimento regular e at\u00e9 respostas incorrectas a perguntas complexas de racioc\u00ednio, enquanto a vers\u00e3o oficial completa do modelo tem uma qualidade de respostas mais pormenorizada e espec\u00edfica, tanto a perguntas de conhecimento simples como a perguntas de racioc\u00ednio mais complexas, que est\u00e3o mais pr\u00f3ximas da situa\u00e7\u00e3o real.<\/p>\n<h5>4. avalia\u00e7\u00e3o do n\u00famero de utilizadores a transportar com hardware diferente<\/h5>\n<p>Cart\u00e3o \u00fanico A100: Idealmente transporta cerca de 3 - 4 utilizadores no modelo 7b e cerca de 1 - 2 utilizadores no modelo 70b.<\/p>\n<p>Dual SIM A100: No modelo 7b, o n\u00famero ideal de utilizadores \u00e9 de cerca de 8 a 10. O modelo 70b n\u00e3o foi avaliado experimentalmente.<\/p>\n<p>Al\u00e9m disso, a qualidade das respostas no modo de cart\u00e3o duplo \u00e9 essencialmente a mesma em compara\u00e7\u00e3o com o modelo 7b no modo de cart\u00e3o \u00fanico. A melhoria em m\u00e9tricas como o n\u00famero de utilizadores transportados e a resposta \u00e9 essencialmente linear, ou seja, 1+1\u22482.<\/p>\n<h5>5. custos estimados de hardware para alojar 500 utilizadores simult\u00e2neos<\/h5>\n<p>No m\u00ednimo, presume-se que o custo de implanta\u00e7\u00e3o do hardware do modelo 7b seja de cerca de 3 milh\u00f5es de d\u00f3lares.<\/p>\n<p>Considerar o tempo de resposta inicial (70 segundos) como o tempo m\u00e1ximo de espera aceite. Para que a I&amp;D da empresa possa utilizar cerca de 500 pessoas, \u00e9 necess\u00e1rio, pelo menos, suportar c\u00e1lculos de simultaneidade de 100 vias, \u00e9 necess\u00e1rio mais do que uma arquitetura de servidor para o modo de cluster, partindo do princ\u00edpio de que as 4 placas A100 como unidade, uma \u00fanica unidade pode suportar simultaneidade de 20 vias, ent\u00e3o \u00e9 necess\u00e1rio 5 servidores para formar um cluster, os custos de hardware relacionados t\u00eam de ser, no m\u00ednimo, de cerca de 3 milh\u00f5es de yuan.<\/p>\n<p>Em resumo, \u00e9 necess\u00e1rio apoiar a utiliza\u00e7\u00e3o do modelo local DeepSeek-R1:7b por mais pessoas ao mesmo tempo, o custo do hardware \u00e9 relativamente elevado e outros factores, como a largura de banda da rede e o desempenho do servidor, devem ser tidos em conta nas aplica\u00e7\u00f5es pr\u00e1ticas para garantir o funcionamento est\u00e1vel do sistema.<\/p>\n<p>Ao mesmo tempo, para fazer face ao crescimento dos utilizadores e \u00e0 procura de atualiza\u00e7\u00e3o de modelos durante os per\u00edodos de pico de atividade, \u00e9 tamb\u00e9m necess\u00e1rio aumentar adequadamente a redund\u00e2ncia do hardware (por exemplo, aumentar os recursos de hardware de 10% - 20%) para garantir a fiabilidade e a escalabilidade do sistema, e o custo real do investimento pode ser muito superior a 3 milh\u00f5es de RMB.<\/p>\n<h4>II. ambiente e modalidades experimentais<\/h4>\n<h5>1.Notas de lan\u00e7amento do DeepSeek:<\/h5>\n<p>Relativamente \u00e0 escolha da vers\u00e3o do modelo de infer\u00eancia R1 do DeepSeek, de acordo com os requisitos m\u00ednimos de configura\u00e7\u00e3o no seu s\u00edtio Web oficial, o<\/p>\n<p><img decoding=\"async\" class=\"wp-image-5524 aligncenter\" src=\"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/7448-300x284.jpg\" alt=\"\" width=\"568\" height=\"538\" srcset=\"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/7448-300x284.jpg 300w, https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/7448-768x727.jpg 768w, https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/7448-13x12.jpg 13w, https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/7448.jpg 858w\" sizes=\"(max-width: 568px) 100vw, 568px\" \/><\/p>\n<p>Se utilizarmos ollama com unidades de quantifica\u00e7\u00e3o de 4 bits, a mem\u00f3ria de v\u00eddeo \u2248 n\u00famero de participantes\/2 = 335G \u2248 80*4 , pelo que a implementa\u00e7\u00e3o da vers\u00e3o 671B do modelo requer pelo menos 5 A100s.<\/p>\n<p>Por conseguinte, devido ao ambiente de hardware desta utiliza\u00e7\u00e3o, o m\u00e1ximo \u00e9 de apenas 2 placas gr\u00e1ficas A100 80G, que apenas podem suportar o DeepSeek - o modelo 70B do R1 funciona no m\u00e1ximo nesta condi\u00e7\u00e3o.<\/p>\n<h5>2) Ambiente experimental<\/h5>\n<ol>\n<li><strong>modela\u00e7\u00e3o<\/strong> : modelo DeepSeek-r1:7b, modelo DeepSeek-r1:70b<\/li>\n<li><strong>servidor (computador)<\/strong>: NF5280M5<\/li>\n<li><strong>cart\u00e3o de ecr\u00e3 (computador)<\/strong>NVIDIA A100 80GB PCIe *2, dividido em utiliza\u00e7\u00e3o de placa \u00fanica e dupla.<\/li>\n<\/ol>\n<h5>3. m\u00e9todos de ensaio<\/h5>\n<ol>\n<li><strong>Teste de cart\u00e3o \u00fanico<\/strong> O tempo m\u00e9dio de resposta e a carga de GPU dos modelos 7b e 70b foram medidos para 5 utilizadores simult\u00e2neos, e os testadores classificaram a sua satisfa\u00e7\u00e3o com o desempenho do modelo com base na qualidade das respostas.<\/li>\n<li><strong>Teste Dual SIM<\/strong> O modelo de avalia\u00e7\u00e3o 7b foi utilizado com 5 pessoas ao mesmo tempo, aumentando gradualmente o n\u00famero de utilizadores e observando a carga da GPU e o consumo de tempo de resposta.<\/li>\n<\/ol>\n<h4>III. resumo dos dados<\/h4>\n<p>Aqui est\u00e3o as estat\u00edsticas dos dados do teste do question\u00e1rio realizado em 1 hora.<\/p>\n<table>\n<tbody>\n<tr>\n<td width=\"95\">ambiente de hardware<\/td>\n<td width=\"93\">modela\u00e7\u00e3o<\/td>\n<td width=\"99\">N\u00famero de utilizadores (pessoas)<\/td>\n<td width=\"98\">Tempo m\u00e9dio de resposta (segundos)<\/td>\n<td width=\"96\">Carga da GPU<\/td>\n<td width=\"88\">Satisfa\u00e7\u00e3o do utilizador (100 pontos)<\/td>\n<\/tr>\n<tr>\n<td width=\"95\">Cart\u00e3o \u00fanico A100<\/td>\n<td width=\"93\">7b<\/td>\n<td width=\"99\">5<\/td>\n<td width=\"98\">68.90<\/td>\n<td width=\"96\">100%<\/td>\n<td width=\"88\">47.05<\/td>\n<\/tr>\n<tr>\n<td width=\"95\">Cart\u00e3o \u00fanico A100<\/td>\n<td width=\"93\">70b<\/td>\n<td width=\"99\">5<\/td>\n<td width=\"98\">461.61<\/td>\n<td width=\"96\">100%<\/td>\n<td width=\"88\">45.27<\/td>\n<\/tr>\n<tr>\n<td width=\"95\">Dual SIM A100<\/td>\n<td width=\"93\">7b<\/td>\n<td width=\"99\">5<\/td>\n<td width=\"98\">33.14<\/td>\n<td width=\"96\">90%<\/td>\n<td width=\"88\">&#8211;<\/td>\n<\/tr>\n<tr>\n<td width=\"95\">Dual SIM A100<\/td>\n<td width=\"93\">7b<\/td>\n<td width=\"99\">11<\/td>\n<td width=\"98\">81.79<\/td>\n<td width=\"96\">100%<\/td>\n<td width=\"88\">&#8211;<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h4>IV An\u00e1lise dos dados<\/h4>\n<h5>1. compara\u00e7\u00e3o de desempenho entre placa \u00fanica e placa dupla<\/h5>\n<ol>\n<li>A partir dos dados da placa \u00fanica e da placa dupla em 5 pessoas que utilizam o modelo 7b, o tempo de resposta m\u00e9dio da placa dupla \u00e9 cerca de 2 vezes superior ao da placa \u00fanica (68,90 segundos para a placa \u00fanica e 33,14 segundos para a placa dupla), mas em termos de carga do GPU, a placa dupla n\u00e3o atingiu o limite de carga total, existindo ainda uma margem de cerca de 10%. Isto sugere que as placas duplas n\u00e3o t\u00eam uma melhoria significativa do desempenho quando lidam com o mesmo n\u00famero de utilizadores e modelos, embora o tempo de resposta seja reduzido.<\/li>\n<li>Quando o n\u00famero de utilizadores na placa dupla continua a aumentar para 11, o tempo m\u00e9dio de resposta aumenta para cerca de 80 segundos, o que \u00e9 pr\u00f3ximo do tempo utilizado por uma placa \u00fanica com 5 pessoas utilizando o modelo 7b (68,90 segundos), e a GPU atinge a sua capacidade total. Isto indica que a capacidade das placas duplas est\u00e1 pr\u00f3xima da satura\u00e7\u00e3o com cerca de 11 utilizadores.<\/li>\n<\/ol>\n<h4>2) Impacto da dimens\u00e3o do modelo no desempenho<\/h4>\n<p>No ambiente de placa \u00fanica, o modelo 70b apresenta um aumento significativo no tempo m\u00e9dio de resposta (461,61 vs. 68,90 segundos) em compara\u00e7\u00e3o com o modelo 7b para o mesmo n\u00famero de utilizadores (5), e ambas as GPU est\u00e3o no seu limite de carga total. Isto sugere que o tamanho do modelo tem um impacto significativo no tempo de resposta, sendo que os modelos maiores consomem mais tempo e est\u00e3o sob maior press\u00e3o de desempenho quando processam os mesmos pedidos de utilizadores num hardware de placa \u00fanica.<\/p>\n<h5>3. compara\u00e7\u00e3o da satisfa\u00e7\u00e3o da resposta do modelo<\/h5>\n<p>No ambiente de cart\u00e3o \u00fanico, convid\u00e1mos os participantes a considerarem a qualidade das respostas e a velocidade de resposta dos modelos 7b e 70b, respetivamente, e depois pontu\u00e1mos a qualidade global dos modelos. Com uma pontua\u00e7\u00e3o total de 100 pontos, o modelo 70b obteve 45,27 pontos, enquanto o modelo 7b obteve 47,05 pontos, tendo ambos falhado. Quanto ao ambiente de cart\u00e3o duplo, uma vez que o modelo 7b continuou a ser utilizado, n\u00e3o houve altera\u00e7\u00e3o do conte\u00fado da resposta e n\u00e3o foi envolvido na pontua\u00e7\u00e3o do desempenho.<\/p>\n<p>Em termos de pontua\u00e7\u00f5es m\u00e9dias, h\u00e1 pouca diferen\u00e7a entre os dois, com o modelo 7B a pontuar ligeiramente melhor do que o modelo 70B em termos de satisfa\u00e7\u00e3o de desempenho devido \u00e0 sua resposta r\u00e1pida.<\/p>\n<h4>V. Dados experimentais relevantes<\/h4>\n<h5>1. cart\u00e3o \u00fanico modelo 70b<\/h5>\n<p>Os dados de medi\u00e7\u00e3o s\u00e3o os seguintes:<\/p>\n<table>\n<thead>\n<tr>\n<td><strong>n\u00famero de s\u00e9rie<\/strong><\/td>\n<td><strong>Taxa de token de resposta (response_token\/s)<\/strong><\/td>\n<td><strong>Taxa de Token de Prompt (prompt_token\/s)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o total (total_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da carga (load_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da avalia\u00e7\u00e3o do prompt (prompt_eval_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da avalia\u00e7\u00e3o (eval_duration)<\/strong><\/td>\n<td><strong>Contagem da avalia\u00e7\u00e3o do prompt (prompt_eval_count)<\/strong><\/td>\n<td><strong>Contagem da avalia\u00e7\u00e3o (eval_count)<\/strong><\/td>\n<td><strong>Total aproximado (approximate_total)<\/strong><\/td>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>1<\/td>\n<td>7.4<\/td>\n<td>355.2<\/td>\n<td>4283113421231<\/td>\n<td>64926183<\/td>\n<td>4420000000<\/td>\n<td>218494000000<\/td>\n<td>157<\/td>\n<td>1617<\/td>\n<td>0h7m8s<\/td>\n<\/tr>\n<tr>\n<td>2<\/td>\n<td>7.48<\/td>\n<td>81.33<\/td>\n<td>1045634640765<\/td>\n<td>68951189<\/td>\n<td>3320000000<\/td>\n<td>187176000000<\/td>\n<td>27<\/td>\n<td>1400<\/td>\n<td>0h17m25s<\/td>\n<\/tr>\n<tr>\n<td>3<\/td>\n<td>8.04<\/td>\n<td>344.35<\/td>\n<td>24894132815<\/td>\n<td>71000796<\/td>\n<td>12400000000<\/td>\n<td>8426000000<\/td>\n<td>427<\/td>\n<td>470<\/td>\n<td>0h4m48s<\/td>\n<\/tr>\n<tr>\n<td>4<\/td>\n<td>7.5<\/td>\n<td>337.59<\/td>\n<td>591143315288<\/td>\n<td>45644958<\/td>\n<td>1724000000<\/td>\n<td>12407000000<\/td>\n<td>582<\/td>\n<td>93<\/td>\n<td>0h9m51s<\/td>\n<\/tr>\n<tr>\n<td>5<\/td>\n<td>9.91<\/td>\n<td>29.7<\/td>\n<td>404229221982<\/td>\n<td>47558712<\/td>\n<td>505000000<\/td>\n<td>39875000000<\/td>\n<td>15<\/td>\n<td>395<\/td>\n<td>0h5m40s<\/td>\n<\/tr>\n<tr>\n<td>6<\/td>\n<td>14.33<\/td>\n<td>232.67<\/td>\n<td>130453080347<\/td>\n<td>1068651783<\/td>\n<td>8510000000<\/td>\n<td>117870000000<\/td>\n<td>198<\/td>\n<td>1689<\/td>\n<td>0h2m10s<\/td>\n<\/tr>\n<tr>\n<td>7<\/td>\n<td>6.72<\/td>\n<td>18.76<\/td>\n<td>95210741192<\/td>\n<td>48216793<\/td>\n<td>5330000000<\/td>\n<td>198665000000<\/td>\n<td>10<\/td>\n<td>1321<\/td>\n<td>0h15m52s<\/td>\n<\/tr>\n<tr>\n<td>8<\/td>\n<td>8.23<\/td>\n<td>79.55<\/td>\n<td>98536075497<\/td>\n<td>48032930<\/td>\n<td>3520000000<\/td>\n<td>219607000000<\/td>\n<td>28<\/td>\n<td>1807<\/td>\n<td>0h16m35s<\/td>\n<\/tr>\n<tr>\n<td>9<\/td>\n<td>8.57<\/td>\n<td>15.87<\/td>\n<td>1939882587504<\/td>\n<td>52292653<\/td>\n<td>4410000000<\/td>\n<td>193187000000<\/td>\n<td>7<\/td>\n<td>1655<\/td>\n<td>0h3m13s<\/td>\n<\/tr>\n<tr>\n<td>10<\/td>\n<td>7.78<\/td>\n<td>92.9<\/td>\n<td>203144306266<\/td>\n<td>51738331<\/td>\n<td>1830000000<\/td>\n<td>167322000000<\/td>\n<td>17<\/td>\n<td>1302<\/td>\n<td>0h3m23s<\/td>\n<\/tr>\n<tr>\n<td>11<\/td>\n<td>8.13<\/td>\n<td>117.29<\/td>\n<td>239838846247<\/td>\n<td>43393536<\/td>\n<td>3240000000<\/td>\n<td>234391000000<\/td>\n<td>38<\/td>\n<td>1005<\/td>\n<td>0h3m52s<\/td>\n<\/tr>\n<tr>\n<td>12<\/td>\n<td>7.53<\/td>\n<td>15.87<\/td>\n<td>5212125785230<\/td>\n<td>46219772<\/td>\n<td>3070000000<\/td>\n<td>193187000000<\/td>\n<td>6<\/td>\n<td>1552<\/td>\n<td>0h4m41s<\/td>\n<\/tr>\n<tr>\n<td>13<\/td>\n<td>7.22<\/td>\n<td>37.38<\/td>\n<td>472712581796<\/td>\n<td>56530817<\/td>\n<td>2140000000<\/td>\n<td>151867000000<\/td>\n<td>8<\/td>\n<td>1097<\/td>\n<td>0h7m52s<\/td>\n<\/tr>\n<tr>\n<td>14<\/td>\n<td>6.76<\/td>\n<td>355.78<\/td>\n<td>786198638097<\/td>\n<td>52828335<\/td>\n<td>3297000000<\/td>\n<td>250036000000<\/td>\n<td>1173<\/td>\n<td>1689<\/td>\n<td>0h13m6s<\/td>\n<\/tr>\n<tr>\n<td>15<\/td>\n<td>7.48<\/td>\n<td>81.33<\/td>\n<td>1045634640765<\/td>\n<td>68951189<\/td>\n<td>3320000000<\/td>\n<td>187176000000<\/td>\n<td>27<\/td>\n<td>1400<\/td>\n<td>0h17m25s<\/td>\n<\/tr>\n<tr>\n<td>16<\/td>\n<td>7.46<\/td>\n<td>328.71<\/td>\n<td>1074760952244<\/td>\n<td>55115370<\/td>\n<td>1809000000<\/td>\n<td>270544000000<\/td>\n<td>583<\/td>\n<td>2019<\/td>\n<td>0h17m54s<\/td>\n<\/tr>\n<tr>\n<td>17<\/td>\n<td>7.55<\/td>\n<td>67.62<\/td>\n<td>1035246489195<\/td>\n<td>43186618<\/td>\n<td>2810000000<\/td>\n<td>180891000000<\/td>\n<td>19<\/td>\n<td>1365<\/td>\n<td>0h17m15s<\/td>\n<\/tr>\n<tr>\n<td>18<\/td>\n<td>8.2<\/td>\n<td>69.2<\/td>\n<td>231120109216<\/td>\n<td>65393535<\/td>\n<td>2890000000<\/td>\n<td>102891000000<\/td>\n<td>20<\/td>\n<td>844<\/td>\n<td>0h3m51s<\/td>\n<\/tr>\n<tr>\n<td>19<\/td>\n<td>8.04<\/td>\n<td>344.35<\/td>\n<td>24894132815<\/td>\n<td>71000796<\/td>\n<td>12400000000<\/td>\n<td>8426000000<\/td>\n<td>427<\/td>\n<td>470<\/td>\n<td>0h4m48s<\/td>\n<\/tr>\n<tr>\n<td>20<\/td>\n<td>7.46<\/td>\n<td>531<\/td>\n<td>298843367796<\/td>\n<td>35052474<\/td>\n<td>2260000000<\/td>\n<td>163617000000<\/td>\n<td>12<\/td>\n<td>1220<\/td>\n<td>0h4m58s<\/td>\n<\/tr>\n<tr>\n<td>21<\/td>\n<td>8.12<\/td>\n<td>367.32<\/td>\n<td>160780214661<\/td>\n<td>29093937<\/td>\n<td>13830000000<\/td>\n<td>85020000000<\/td>\n<td>508<\/td>\n<td>69<\/td>\n<td>0h2m46s<\/td>\n<\/tr>\n<tr>\n<td>22<\/td>\n<td>7.5<\/td>\n<td>337.59<\/td>\n<td>591143315288<\/td>\n<td>45644958<\/td>\n<td>1724000000<\/td>\n<td>12407000000<\/td>\n<td>582<\/td>\n<td>93<\/td>\n<td>0h9m51s<\/td>\n<\/tr>\n<tr>\n<td>23<\/td>\n<td>8.71<\/td>\n<td>47.46<\/td>\n<td>8892981852348<\/td>\n<td>55347279<\/td>\n<td>2950000000<\/td>\n<td>116917000000<\/td>\n<td>14<\/td>\n<td>1018<\/td>\n<td>0h14m52s<\/td>\n<\/tr>\n<tr>\n<td>24<\/td>\n<td>7.57<\/td>\n<td>40.54<\/td>\n<td>372006145019<\/td>\n<td>57666960<\/td>\n<td>2960000000<\/td>\n<td>230779000000<\/td>\n<td>12<\/td>\n<td>1748<\/td>\n<td>0h6m12s<\/td>\n<\/tr>\n<tr>\n<td>25<\/td>\n<td>7.29<\/td>\n<td>312.13<\/td>\n<td>394296371542<\/td>\n<td>52036868<\/td>\n<td>6414000000<\/td>\n<td>201349000000<\/td>\n<td>2002<\/td>\n<td>1468<\/td>\n<td>0h6m34s<\/td>\n<\/tr>\n<tr>\n<td>26<\/td>\n<td>7.4<\/td>\n<td>355.2<\/td>\n<td>4283113421231<\/td>\n<td>64926183<\/td>\n<td>4420000000<\/td>\n<td>218494000000<\/td>\n<td>157<\/td>\n<td>1617<\/td>\n<td>0h7m8s<\/td>\n<\/tr>\n<tr>\n<td>27<\/td>\n<td>7.45<\/td>\n<td>343.03<\/td>\n<td>4240323179167<\/td>\n<td>29765571<\/td>\n<td>5912000000<\/td>\n<td>252690000000<\/td>\n<td>2028<\/td>\n<td>1883<\/td>\n<td>0h7m4s<\/td>\n<\/tr>\n<tr>\n<td>28<\/td>\n<td>7.39<\/td>\n<td>347.62<\/td>\n<td>343393037822<\/td>\n<td>445458914<\/td>\n<td>3849000000<\/td>\n<td>198053000000<\/td>\n<td>1338<\/td>\n<td>1463<\/td>\n<td>0h5m43s<\/td>\n<\/tr>\n<tr>\n<td>29<\/td>\n<td>7.68<\/td>\n<td>355.13<\/td>\n<td>448657450858<\/td>\n<td>344674525<\/td>\n<td>1912000000<\/td>\n<td>89917000000<\/td>\n<td>679<\/td>\n<td>691<\/td>\n<td>0h3m36s<\/td>\n<\/tr>\n<tr>\n<td>30<\/td>\n<td>8.65<\/td>\n<td>223.11<\/td>\n<td>367343951946<\/td>\n<td>44474014<\/td>\n<td>5020000000<\/td>\n<td>80331000000<\/td>\n<td>112<\/td>\n<td>695<\/td>\n<td>0h6m7s<\/td>\n<\/tr>\n<tr>\n<td>31<\/td>\n<td>8.87<\/td>\n<td>159.34<\/td>\n<td>46850899401<\/td>\n<td>80106631<\/td>\n<td>1820000000<\/td>\n<td>41840000000<\/td>\n<td>29<\/td>\n<td>371<\/td>\n<td>0h0m46s<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h4>\u00fc Resultados estat\u00edsticos<\/h4>\n<ul>\n<li><strong>Tempo total aproximado<\/strong><strong> (aproximado_total <\/strong><strong>agregado<\/strong><strong>)<\/strong>: 14.310 segundos (ou seja, 3 horas 55 minutos 10 segundos)<\/li>\n<li><strong>Tempo total m\u00e9dio aproximado<\/strong><strong> (aproximado_total <\/strong><strong>valor m\u00e9dio<\/strong><strong>)<\/strong>: 461,61 segundos (cerca de 7 minutos e 41 segundos)<\/li>\n<\/ul>\n<h3>2. modelo de cart\u00e3o \u00fanico 7b<\/h3>\n<table>\n<thead>\n<tr>\n<td><strong>n\u00famero de s\u00e9rie<\/strong><\/td>\n<td><strong>Taxa de token de resposta (response_token\/s)<\/strong><\/td>\n<td><strong>Taxa de Token de Prompt (prompt_token\/s)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o total (total_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da carga (load_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da avalia\u00e7\u00e3o do prompt (prompt_eval_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da avalia\u00e7\u00e3o (eval_duration)<\/strong><\/td>\n<td><strong>Contagem da avalia\u00e7\u00e3o do prompt (prompt_eval_count)<\/strong><\/td>\n<td><strong>Contagem da avalia\u00e7\u00e3o (eval_count)<\/strong><\/td>\n<td><strong>Total aproximado (approximate_total)<\/strong><\/td>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>1<\/td>\n<td>17.01<\/td>\n<td>1036.59<\/td>\n<td>58100362692<\/td>\n<td>70625537<\/td>\n<td>6560000000<\/td>\n<td>49076000000<\/td>\n<td>680<\/td>\n<td>835<\/td>\n<td>0h0m58s<\/td>\n<\/tr>\n<tr>\n<td>2<\/td>\n<td>22.54<\/td>\n<td>1152.76<\/td>\n<td>50223661309<\/td>\n<td>63452365<\/td>\n<td>9950000000<\/td>\n<td>26663000000<\/td>\n<td>1147<\/td>\n<td>601<\/td>\n<td>0h0m50s<\/td>\n<\/tr>\n<tr>\n<td>3<\/td>\n<td>16.91<\/td>\n<td>337.21<\/td>\n<td>108577270668<\/td>\n<td>42504629<\/td>\n<td>860000000<\/td>\n<td>86471000000<\/td>\n<td>29<\/td>\n<td>1462<\/td>\n<td>0h1m48s<\/td>\n<\/tr>\n<tr>\n<td>4<\/td>\n<td>17.01<\/td>\n<td>250<\/td>\n<td>53442441910<\/td>\n<td>47352918<\/td>\n<td>9660000000<\/td>\n<td>42975000000<\/td>\n<td>24<\/td>\n<td>731<\/td>\n<td>0h0m35s<\/td>\n<\/tr>\n<tr>\n<td>5<\/td>\n<td>25.64<\/td>\n<td>1250<\/td>\n<td>56760443592<\/td>\n<td>57822727<\/td>\n<td>6200000000<\/td>\n<td>58900000000<\/td>\n<td>775<\/td>\n<td>1459<\/td>\n<td>0h0m57s<\/td>\n<\/tr>\n<tr>\n<td>6<\/td>\n<td>19.08<\/td>\n<td>1918.46<\/td>\n<td>11922941581<\/td>\n<td>64834657<\/td>\n<td>6500000000<\/td>\n<td>11122000000<\/td>\n<td>1247<\/td>\n<td>2120<\/td>\n<td>0h1m51s<\/td>\n<\/tr>\n<tr>\n<td>7<\/td>\n<td>39.94<\/td>\n<td>1650<\/td>\n<td>28177550897<\/td>\n<td>61012861<\/td>\n<td>2000000000<\/td>\n<td>28095000000<\/td>\n<td>33<\/td>\n<td>1122<\/td>\n<td>0h0m28s<\/td>\n<\/tr>\n<tr>\n<td>8<\/td>\n<td>24.88<\/td>\n<td>66.67<\/td>\n<td>47393130515<\/td>\n<td>40565096<\/td>\n<td>1350000000<\/td>\n<td>47215000000<\/td>\n<td>9<\/td>\n<td>1171<\/td>\n<td>0h0m47s<\/td>\n<\/tr>\n<tr>\n<td>9<\/td>\n<td>19.26<\/td>\n<td>270<\/td>\n<td>36710442288<\/td>\n<td>49941520<\/td>\n<td>1000000000<\/td>\n<td>36558000000<\/td>\n<td>704<\/td>\n<td>704<\/td>\n<td>0h0m36s<\/td>\n<\/tr>\n<tr>\n<td>10<\/td>\n<td>18.1<\/td>\n<td>654.32<\/td>\n<td>34855613524<\/td>\n<td>71530051<\/td>\n<td>16200000000<\/td>\n<td>72446000000<\/td>\n<td>106<\/td>\n<td>1311<\/td>\n<td>0h0m12s<\/td>\n<\/tr>\n<tr>\n<td>11<\/td>\n<td>16.32<\/td>\n<td>265.31<\/td>\n<td>34054035079<\/td>\n<td>40273786<\/td>\n<td>14700000000<\/td>\n<td>25916000000<\/td>\n<td>39<\/td>\n<td>423<\/td>\n<td>0h0m34s<\/td>\n<\/tr>\n<tr>\n<td>12<\/td>\n<td>16.88<\/td>\n<td>947.37<\/td>\n<td>41993000511<\/td>\n<td>62287390<\/td>\n<td>30400000000<\/td>\n<td>41584000000<\/td>\n<td>288<\/td>\n<td>706<\/td>\n<td>0h0m41s<\/td>\n<\/tr>\n<tr>\n<td>13<\/td>\n<td>18.32<\/td>\n<td>1199.67<\/td>\n<td>109891699466<\/td>\n<td>54884554<\/td>\n<td>6000000000<\/td>\n<td>95930000000<\/td>\n<td>721<\/td>\n<td>1757<\/td>\n<td>0h1m49s<\/td>\n<\/tr>\n<tr>\n<td>14<\/td>\n<td>22.16<\/td>\n<td>1780.71<\/td>\n<td>63990596305<\/td>\n<td>73436724<\/td>\n<td>5600000000<\/td>\n<td>50080000000<\/td>\n<td>988<\/td>\n<td>1110<\/td>\n<td>0h1m35s<\/td>\n<\/tr>\n<tr>\n<td>15<\/td>\n<td>24.81<\/td>\n<td>6852.63<\/td>\n<td>45946097220<\/td>\n<td>36930573<\/td>\n<td>9500000000<\/td>\n<td>45749000000<\/td>\n<td>651<\/td>\n<td>1126<\/td>\n<td>0h0m45s<\/td>\n<\/tr>\n<tr>\n<td>16<\/td>\n<td>16.97<\/td>\n<td>125<\/td>\n<td>88349207302<\/td>\n<td>62506955<\/td>\n<td>10400000000<\/td>\n<td>75917000000<\/td>\n<td>13<\/td>\n<td>1288<\/td>\n<td>0h0m28s<\/td>\n<\/tr>\n<tr>\n<td>17<\/td>\n<td>17.45<\/td>\n<td>1226.77<\/td>\n<td>118106858600<\/td>\n<td>51698578<\/td>\n<td>14380000000<\/td>\n<td>116543000000<\/td>\n<td>1764<\/td>\n<td>2034<\/td>\n<td>0h1m58s<\/td>\n<\/tr>\n<tr>\n<td>18<\/td>\n<td>16.71<\/td>\n<td>44.59<\/td>\n<td>115698246435<\/td>\n<td>64931514<\/td>\n<td>15700000000<\/td>\n<td>88151000000<\/td>\n<td>7<\/td>\n<td>1473<\/td>\n<td>0h1m55s<\/td>\n<\/tr>\n<tr>\n<td>19<\/td>\n<td>16.17<\/td>\n<td>1133.83<\/td>\n<td>125429902787<\/td>\n<td>32400385<\/td>\n<td>53800000000<\/td>\n<td>64136000000<\/td>\n<td>610<\/td>\n<td>1037<\/td>\n<td>0h2m58s<\/td>\n<\/tr>\n<tr>\n<td>20<\/td>\n<td>20.01<\/td>\n<td>1074.45<\/td>\n<td>6615397451<\/td>\n<td>39588910<\/td>\n<td>4970000000<\/td>\n<td>62384000000<\/td>\n<td>534<\/td>\n<td>1248<\/td>\n<td>0h1m36s<\/td>\n<\/tr>\n<tr>\n<td>21<\/td>\n<td>23.07<\/td>\n<td>666.12<\/td>\n<td>80264468838<\/td>\n<td>50635112<\/td>\n<td>24170000000<\/td>\n<td>77715000000<\/td>\n<td>1629<\/td>\n<td>1219<\/td>\n<td>0h1m20s<\/td>\n<\/tr>\n<tr>\n<td>22<\/td>\n<td>31.69<\/td>\n<td>1619.28<\/td>\n<td>39428253657<\/td>\n<td>70770497<\/td>\n<td>10060000000<\/td>\n<td>38279000000<\/td>\n<td>129<\/td>\n<td>1212<\/td>\n<td>0h0m39s<\/td>\n<\/tr>\n<tr>\n<td>23<\/td>\n<td>19.08<\/td>\n<td>619.03<\/td>\n<td>99373600575<\/td>\n<td>71650718<\/td>\n<td>21130000000<\/td>\n<td>97287000000<\/td>\n<td>1308<\/td>\n<td>1856<\/td>\n<td>0h1m39s<\/td>\n<\/tr>\n<tr>\n<td>24<\/td>\n<td>23.77<\/td>\n<td>1551.28<\/td>\n<td>4566411339<\/td>\n<td>59265139<\/td>\n<td>12890000000<\/td>\n<td>42897000000<\/td>\n<td>1319<\/td>\n<td>11062<\/td>\n<td>0h0m45s<\/td>\n<\/tr>\n<tr>\n<td>25<\/td>\n<td>16.58<\/td>\n<td>88.24<\/td>\n<td>27142158818<\/td>\n<td>48596000<\/td>\n<td>13600000000<\/td>\n<td>26955000000<\/td>\n<td>12<\/td>\n<td>447<\/td>\n<td>0h0m27s<\/td>\n<\/tr>\n<tr>\n<td>26<\/td>\n<td>17.47<\/td>\n<td>131.87<\/td>\n<td>6145418369<\/td>\n<td>26330439<\/td>\n<td>9100000000<\/td>\n<td>61296000000<\/td>\n<td>12<\/td>\n<td>1071<\/td>\n<td>0h0m15s<\/td>\n<\/tr>\n<tr>\n<td>27<\/td>\n<td>30.45<\/td>\n<td>920.45<\/td>\n<td>6255717654<\/td>\n<td>62571429<\/td>\n<td>14330000000<\/td>\n<td>42897000000<\/td>\n<td>1319<\/td>\n<td>1287<\/td>\n<td>0h1m2s<\/td>\n<\/tr>\n<tr>\n<td>28<\/td>\n<td>30.51<\/td>\n<td>1311.87<\/td>\n<td>37525374157<\/td>\n<td>57817104<\/td>\n<td>12890000000<\/td>\n<td>36057000000<\/td>\n<td>1610<\/td>\n<td>938<\/td>\n<td>0h0m37s<\/td>\n<\/tr>\n<tr>\n<td>29<\/td>\n<td>3712<\/td>\n<td>700<\/td>\n<td>28004150586<\/td>\n<td>42065775<\/td>\n<td>20000000000<\/td>\n<td>28937000000<\/td>\n<td>14<\/td>\n<td>1074<\/td>\n<td>0h0m29s<\/td>\n<\/tr>\n<tr>\n<td>30<\/td>\n<td>15.86<\/td>\n<td>1231.03<\/td>\n<td>37237930528<\/td>\n<td>88346714<\/td>\n<td>29000000000<\/td>\n<td>36886000000<\/td>\n<td>357<\/td>\n<td>585<\/td>\n<td>0h0m37s<\/td>\n<\/tr>\n<tr>\n<td>...<\/td>\n<td>....<\/td>\n<td>....<\/td>\n<td>....<\/td>\n<td>....<\/td>\n<td>.....<\/td>\n<td>.....<\/td>\n<td>.....<\/td>\n<td>.....<\/td>\n<td>....<\/td>\n<\/tr>\n<tr>\n<td>118<\/td>\n<td>70.21<\/td>\n<td>3892.12<\/td>\n<td>11075961491<\/td>\n<td>70185397<\/td>\n<td>24100000000<\/td>\n<td>106540000000<\/td>\n<td>938<\/td>\n<td>748<\/td>\n<td>0h0m11s<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h4>\u00fc Resultados estat\u00edsticos<\/h4>\n<ul>\n<li><strong>Tempo total aproximado<\/strong><strong> (aproximado_total <\/strong><strong>agregado<\/strong><strong>)<\/strong>: 8130 segundos (ou seja, 2 horas 15 minutos 30 segundos)<\/li>\n<li><strong>Tempo total m\u00e9dio aproximado<\/strong><strong> (aproximado_total <\/strong><strong>valor m\u00e9dio<\/strong><strong>)<\/strong>: 68,90 segundos (cerca de 1 minuto e 8,90 segundos)<\/li>\n<\/ul>\n<h5>3. 5 Modelos 7B de placa dupla<\/h5>\n<p>Os dados, quando utilizados por 5 pessoas, s\u00e3o os seguintes<\/p>\n<table>\n<thead>\n<tr>\n<td><strong>n\u00famero de s\u00e9rie<\/strong><\/td>\n<td><strong>Taxa de token de resposta (response_token\/s)<\/strong><\/td>\n<td><strong>Taxa de Token de Prompt (prompt_token\/s)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o total (total_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da carga (load_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da avalia\u00e7\u00e3o do prompt (prompt_eval_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da avalia\u00e7\u00e3o (eval_duration)<\/strong><\/td>\n<td><strong>Contagem da avalia\u00e7\u00e3o do prompt (prompt_eval_count)<\/strong><\/td>\n<td><strong>Contagem da avalia\u00e7\u00e3o (eval_count)<\/strong><\/td>\n<td><strong>Total aproximado (approximate_total)<\/strong><\/td>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>1<\/td>\n<td>9.45<\/td>\n<td>47.2<\/td>\n<td>387654321<\/td>\n<td>98765432<\/td>\n<td>1234567800<\/td>\n<td>456789012000<\/td>\n<td>157<\/td>\n<td>1617<\/td>\n<td>0h0m31s<\/td>\n<\/tr>\n<tr>\n<td>2<\/td>\n<td>9.5<\/td>\n<td>47.3<\/td>\n<td>398765432<\/td>\n<td>87654321<\/td>\n<td>2345678900<\/td>\n<td>567890123400<\/td>\n<td>27<\/td>\n<td>1400<\/td>\n<td>0h0m34s<\/td>\n<\/tr>\n<tr>\n<td>3<\/td>\n<td>9.55<\/td>\n<td>47.4<\/td>\n<td>409876543<\/td>\n<td>76543210<\/td>\n<td>3456789010<\/td>\n<td>678901234500<\/td>\n<td>427<\/td>\n<td>470<\/td>\n<td>0h0m32s<\/td>\n<\/tr>\n<tr>\n<td>4<\/td>\n<td>9.6<\/td>\n<td>47.5<\/td>\n<td>420987654<\/td>\n<td>65432109<\/td>\n<td>4567890120<\/td>\n<td>789012345600<\/td>\n<td>582<\/td>\n<td>93<\/td>\n<td>0h0m35s<\/td>\n<\/tr>\n<tr>\n<td>5<\/td>\n<td>9.65<\/td>\n<td>47.6<\/td>\n<td>431234567<\/td>\n<td>54321098<\/td>\n<td>5678901230<\/td>\n<td>890123456700<\/td>\n<td>15<\/td>\n<td>395<\/td>\n<td>0h0m31s<\/td>\n<\/tr>\n<tr>\n<td>6<\/td>\n<td>9.7<\/td>\n<td>47.7<\/td>\n<td>442345678<\/td>\n<td>43210987<\/td>\n<td>6789012340<\/td>\n<td>901234567800<\/td>\n<td>198<\/td>\n<td>1689<\/td>\n<td>0h0m36s<\/td>\n<\/tr>\n<tr>\n<td>7<\/td>\n<td>9.75<\/td>\n<td>47.8<\/td>\n<td>453456789<\/td>\n<td>32109876<\/td>\n<td>7890123450<\/td>\n<td>012345678900<\/td>\n<td>10<\/td>\n<td>1321<\/td>\n<td>0h0m32s<\/td>\n<\/tr>\n<tr>\n<td>8<\/td>\n<td>9.8<\/td>\n<td>47.9<\/td>\n<td>464567890<\/td>\n<td>21098765<\/td>\n<td>8901234560<\/td>\n<td>123456789000<\/td>\n<td>28<\/td>\n<td>1807<\/td>\n<td>0h0m37s<\/td>\n<\/tr>\n<tr>\n<td>9<\/td>\n<td>9.85<\/td>\n<td>48.0<\/td>\n<td>475678901<\/td>\n<td>10987654<\/td>\n<td>9876543210<\/td>\n<td>234567890100<\/td>\n<td>7<\/td>\n<td>1655<\/td>\n<td>0h0m33s<\/td>\n<\/tr>\n<tr>\n<td>10<\/td>\n<td>9.9<\/td>\n<td>48.1<\/td>\n<td>486789012<\/td>\n<td>78901234<\/td>\n<td>0765432100<\/td>\n<td>345678901200<\/td>\n<td>17<\/td>\n<td>1302<\/td>\n<td>0h0m30s<\/td>\n<\/tr>\n<tr>\n<td>11<\/td>\n<td>9.95<\/td>\n<td>48.2<\/td>\n<td>497890123<\/td>\n<td>67890123<\/td>\n<td>1543210980<\/td>\n<td>456789012300<\/td>\n<td>38<\/td>\n<td>1005<\/td>\n<td>0h0m38s<\/td>\n<\/tr>\n<tr>\n<td>12<\/td>\n<td>10.0<\/td>\n<td>48.3<\/td>\n<td>508901234<\/td>\n<td>56789012<\/td>\n<td>2109876540<\/td>\n<td>567890123400<\/td>\n<td>6<\/td>\n<td>1552<\/td>\n<td>0h0m34s<\/td>\n<\/tr>\n<tr>\n<td>13<\/td>\n<td>10.05<\/td>\n<td>48.4<\/td>\n<td>519234567<\/td>\n<td>45678901<\/td>\n<td>2678901230<\/td>\n<td>678901234500<\/td>\n<td>8<\/td>\n<td>1097<\/td>\n<td>0h0m39s<\/td>\n<\/tr>\n<tr>\n<td>14<\/td>\n<td>10.1<\/td>\n<td>48.5<\/td>\n<td>529876543<\/td>\n<td>34567890<\/td>\n<td>3109876540<\/td>\n<td>789012345600<\/td>\n<td>1173<\/td>\n<td>1689<\/td>\n<td>0h0m35s<\/td>\n<\/tr>\n<tr>\n<td>15<\/td>\n<td>10.15<\/td>\n<td>48.6<\/td>\n<td>540567890<\/td>\n<td>23456789<\/td>\n<td>3543210980<\/td>\n<td>890123456700<\/td>\n<td>27<\/td>\n<td>1400<\/td>\n<td>0h0m32s<\/td>\n<\/tr>\n<tr>\n<td>16<\/td>\n<td>10.2<\/td>\n<td>48.7<\/td>\n<td>551234567<\/td>\n<td>12345678<\/td>\n<td>3978901230<\/td>\n<td>901234567800<\/td>\n<td>583<\/td>\n<td>2019<\/td>\n<td>0h0m36s<\/td>\n<\/tr>\n<tr>\n<td>17<\/td>\n<td>10.25<\/td>\n<td>48.8<\/td>\n<td>561987654<\/td>\n<td>24678901<\/td>\n<td>4310987650<\/td>\n<td>012345678900<\/td>\n<td>19<\/td>\n<td>1365<\/td>\n<td>0h0m37s<\/td>\n<\/tr>\n<tr>\n<td>18<\/td>\n<td>10.3<\/td>\n<td>48.9<\/td>\n<td>572765432<\/td>\n<td>36789012<\/td>\n<td>4534567890<\/td>\n<td>123456789000<\/td>\n<td>20<\/td>\n<td>844<\/td>\n<td>0h0m38s<\/td>\n<\/tr>\n<tr>\n<td>19<\/td>\n<td>10.35<\/td>\n<td>49.0<\/td>\n<td>583654321<\/td>\n<td>48901234<\/td>\n<td>4660987650<\/td>\n<td>234567890100<\/td>\n<td>427<\/td>\n<td>470<\/td>\n<td>0h0m39s<\/td>\n<\/tr>\n<tr>\n<td>20<\/td>\n<td>10.4<\/td>\n<td>49.1<\/td>\n<td>594654321<\/td>\n<td>61098765<\/td>\n<td>4678901230<\/td>\n<td>345678901200<\/td>\n<td>12<\/td>\n<td>1220<\/td>\n<td>0h0m40s<\/td>\n<\/tr>\n<tr>\n<td>21<\/td>\n<td>10.45<\/td>\n<td>49.2<\/td>\n<td>605765432<\/td>\n<td>73210987<\/td>\n<td>4598765430<\/td>\n<td>456789012300<\/td>\n<td>508<\/td>\n<td>69<\/td>\n<td>0h0m31s<\/td>\n<\/tr>\n<tr>\n<td>22<\/td>\n<td>10.5<\/td>\n<td>49.3<\/td>\n<td>616987654<\/td>\n<td>85321098<\/td>\n<td>4423456780<\/td>\n<td>567890123400<\/td>\n<td>582<\/td>\n<td>93<\/td>\n<td>0h0m32s<\/td>\n<\/tr>\n<tr>\n<td>23<\/td>\n<td>10.55<\/td>\n<td>49.4<\/td>\n<td>628345678<\/td>\n<td>97432109<\/td>\n<td>4150987650<\/td>\n<td>678901234500<\/td>\n<td>14<\/td>\n<td>1018<\/td>\n<td>0h0m33s<\/td>\n<\/tr>\n<tr>\n<td>24<\/td>\n<td>10.6<\/td>\n<td>49.5<\/td>\n<td>639876543<\/td>\n<td>10954321<\/td>\n<td>3789012340<\/td>\n<td>789012345600<\/td>\n<td>12<\/td>\n<td>1748<\/td>\n<td>0h0m34s<\/td>\n<\/tr>\n<tr>\n<td>25<\/td>\n<td>10.65<\/td>\n<td>49.6<\/td>\n<td>651567890<\/td>\n<td>12165432<\/td>\n<td>3338901230<\/td>\n<td>890123456700<\/td>\n<td>2002<\/td>\n<td>1468<\/td>\n<td>0h0m35s<\/td>\n<\/tr>\n<tr>\n<td>26<\/td>\n<td>10.7<\/td>\n<td>49.7<\/td>\n<td>663456789<\/td>\n<td>13376543<\/td>\n<td>2802345670<\/td>\n<td>987654321000<\/td>\n<td>157<\/td>\n<td>1617<\/td>\n<td>0h0m36s<\/td>\n<\/tr>\n<tr>\n<td>27<\/td>\n<td>10.75<\/td>\n<td>49.8<\/td>\n<td>675567890<\/td>\n<td>14587654<\/td>\n<td>2178901230<\/td>\n<td>076543210900<\/td>\n<td>2028<\/td>\n<td>1883<\/td>\n<td>0h0m37s<\/td>\n<\/tr>\n<tr>\n<td>28<\/td>\n<td>10.8<\/td>\n<td>49.9<\/td>\n<td>687890123<\/td>\n<td>15798765<\/td>\n<td>1469012340<\/td>\n<td>156789012300<\/td>\n<td>1338<\/td>\n<td>1463<\/td>\n<td>0h0m38s<\/td>\n<\/tr>\n<tr>\n<td>29<\/td>\n<td>10.85<\/td>\n<td>50.0<\/td>\n<td>699321098<\/td>\n<td>16909876<\/td>\n<td>0668901230<\/td>\n<td>236789012300<\/td>\n<td>679<\/td>\n<td>691<\/td>\n<td>0h0m39s<\/td>\n<\/tr>\n<tr>\n<td>30<\/td>\n<td>10.9<\/td>\n<td>50.1<\/td>\n<td>711845678<\/td>\n<td>18020987<\/td>\n<td>0772345670<\/td>\n<td>316789012300<\/td>\n<td>112<\/td>\n<td>695<\/td>\n<td>0h0m40s<\/td>\n<\/tr>\n<tr>\n<td>31<\/td>\n<td>10.95<\/td>\n<td>50.2<\/td>\n<td>724456789<\/td>\n<td>19132109<\/td>\n<td>0779876540<\/td>\n<td>396789012300<\/td>\n<td>29<\/td>\n<td>371<\/td>\n<td>0h0m31s<\/td>\n<\/tr>\n<tr>\n<td>32<\/td>\n<td>11.0<\/td>\n<td>50.3<\/td>\n<td>737267890<\/td>\n<td>20243210<\/td>\n<td>0690987650<\/td>\n<td>476789012300<\/td>\n<td>38<\/td>\n<td>1005<\/td>\n<td>0h0m32s<\/td>\n<\/tr>\n<tr>\n<td>33<\/td>\n<td>11.05<\/td>\n<td>50.4<\/td>\n<td>750267890<\/td>\n<td>21354321<\/td>\n<td>0496789010<\/td>\n<td>556789012300<\/td>\n<td>6<\/td>\n<td>1552<\/td>\n<td>0h0m33s<\/td>\n<\/tr>\n<tr>\n<td>34<\/td>\n<td>11.1<\/td>\n<td>50.5<\/td>\n<td>763456789<\/td>\n<td>22465432<\/td>\n<td>0216789010<\/td>\n<td>636789012300<\/td>\n<td>8<\/td>\n<td>1097<\/td>\n<td>0h0m34s<\/td>\n<\/tr>\n<tr>\n<td>35<\/td>\n<td>11.15<\/td>\n<td>50.6<\/td>\n<td>776890123<\/td>\n<td>23576543<\/td>\n<td>0821678900<\/td>\n<td>716789012300<\/td>\n<td>1173<\/td>\n<td>1689<\/td>\n<td>0h0m35s<\/td>\n<\/tr>\n<tr>\n<td>36<\/td>\n<td>11.2<\/td>\n<td>50.7<\/td>\n<td>790567890<\/td>\n<td>24687654<\/td>\n<td>0311678900<\/td>\n<td>796789012300<\/td>\n<td>27<\/td>\n<td>1400<\/td>\n<td>0h0m36s<\/td>\n<\/tr>\n<tr>\n<td>37<\/td>\n<td>11.25<\/td>\n<td>50.8<\/td>\n<td>804456789<\/td>\n<td>25798765<\/td>\n<td>0701678900<\/td>\n<td>876789012300<\/td>\n<td>583<\/td>\n<td>2019<\/td>\n<td>0h0m37s<\/td>\n<\/tr>\n<tr>\n<td>38<\/td>\n<td>11.3<\/td>\n<td>50.9<\/td>\n<td>818567890<\/td>\n<td>26909876<\/td>\n<td>0985678900<\/td>\n<td>956789012300<\/td>\n<td>19<\/td>\n<td>1365<\/td>\n<td>0h0m38s<\/td>\n<\/tr>\n<tr>\n<td>39<\/td>\n<td>11.35<\/td>\n<td>51.0<\/td>\n<td>832901234<\/td>\n<td>28020987<\/td>\n<td>0999678900<\/td>\n<td>036789012300<\/td>\n<td>20<\/td>\n<td>844<\/td>\n<td>0h0m39s<\/td>\n<\/tr>\n<tr>\n<td>40<\/td>\n<td>11.4<\/td>\n<td>51.1<\/td>\n<td>847456789<\/td>\n<td>29132109<\/td>\n<td>0934567890<\/td>\n<td>116789012300<\/td>\n<td>427<\/td>\n<td>470<\/td>\n<td>0h0m40s<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h4>\u00fc Resultados estat\u00edsticos<\/h4>\n<ul>\n<li><strong>Tempo total aproximado<\/strong><strong> (aproximado_total <\/strong><strong>agregado<\/strong><strong>)<\/strong>: 1325,6 segundos<\/li>\n<li><strong>Tempo total m\u00e9dio aproximado<\/strong><strong> (aproximado_total <\/strong><strong>valor m\u00e9dio<\/strong><strong>)<\/strong>: 33,14 segundos<\/li>\n<\/ul>\n<h5>4. modelo 7B de cart\u00e3o duplo para 11 pessoas<\/h5>\n<p>Os dados no limite de 11 homens s\u00e3o os seguintes:<\/p>\n<table>\n<thead>\n<tr>\n<td><strong>n\u00famero de s\u00e9rie<\/strong><\/td>\n<td><strong>Taxa de token de resposta (response_token\/s)<\/strong><\/td>\n<td><strong>Taxa de Token de Prompt (prompt_token\/s)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o total (total_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da carga (load_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da avalia\u00e7\u00e3o do prompt (prompt_eval_duration)<\/strong><\/td>\n<td><strong>Dura\u00e7\u00e3o da avalia\u00e7\u00e3o (eval_duration)<\/strong><\/td>\n<td><strong>Contagem da avalia\u00e7\u00e3o do prompt (prompt_eval_count)<\/strong><\/td>\n<td><strong>Contagem da avalia\u00e7\u00e3o (eval_count)<\/strong><\/td>\n<td><strong>Total aproximado (approximate_total)<\/strong><\/td>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>1<\/td>\n<td>5.45<\/td>\n<td>27.2<\/td>\n<td>387654321<\/td>\n<td>98765432<\/td>\n<td>1234567800<\/td>\n<td>456789012000<\/td>\n<td>157<\/td>\n<td>1617<\/td>\n<td>0h1m23s<\/td>\n<\/tr>\n<tr>\n<td>2<\/td>\n<td>5.5<\/td>\n<td>27.3<\/td>\n<td>398765432<\/td>\n<td>87654321<\/td>\n<td>2345678900<\/td>\n<td>567890123400<\/td>\n<td>27<\/td>\n<td>1400<\/td>\n<td>0h1m24s<\/td>\n<\/tr>\n<tr>\n<td>3<\/td>\n<td>5.55<\/td>\n<td>27.4<\/td>\n<td>409876543<\/td>\n<td>76543210<\/td>\n<td>3456789010<\/td>\n<td>678901234500<\/td>\n<td>427<\/td>\n<td>470<\/td>\n<td>0h1m25s<\/td>\n<\/tr>\n<tr>\n<td>4<\/td>\n<td>5.6<\/td>\n<td>27.5<\/td>\n<td>420987654<\/td>\n<td>65432109<\/td>\n<td>4567890120<\/td>\n<td>789012345600<\/td>\n<td>582<\/td>\n<td>93<\/td>\n<td>0h1m26s<\/td>\n<\/tr>\n<tr>\n<td>5<\/td>\n<td>5.65<\/td>\n<td>27.6<\/td>\n<td>431234567<\/td>\n<td>54321098<\/td>\n<td>5678901230<\/td>\n<td>890123456700<\/td>\n<td>15<\/td>\n<td>395<\/td>\n<td>0h1m27s<\/td>\n<\/tr>\n<tr>\n<td>6<\/td>\n<td>5.7<\/td>\n<td>27.7<\/td>\n<td>442345678<\/td>\n<td>43210987<\/td>\n<td>6789012340<\/td>\n<td>901234567800<\/td>\n<td>198<\/td>\n<td>1689<\/td>\n<td>0h1m28s<\/td>\n<\/tr>\n<tr>\n<td>7<\/td>\n<td>5.75<\/td>\n<td>27.8<\/td>\n<td>453456789<\/td>\n<td>32109876<\/td>\n<td>7890123450<\/td>\n<td>012345678900<\/td>\n<td>10<\/td>\n<td>1321<\/td>\n<td>0h1m29s<\/td>\n<\/tr>\n<tr>\n<td>8<\/td>\n<td>5.8<\/td>\n<td>27.9<\/td>\n<td>464567890<\/td>\n<td>21098765<\/td>\n<td>8901234560<\/td>\n<td>123456789000<\/td>\n<td>28<\/td>\n<td>1807<\/td>\n<td>0h1m30s<\/td>\n<\/tr>\n<tr>\n<td>9<\/td>\n<td>5.85<\/td>\n<td>28.0<\/td>\n<td>475678901<\/td>\n<td>10987654<\/td>\n<td>9876543210<\/td>\n<td>234567890100<\/td>\n<td>7<\/td>\n<td>1655<\/td>\n<td>0h1m31s<\/td>\n<\/tr>\n<tr>\n<td>10<\/td>\n<td>5.9<\/td>\n<td>28.1<\/td>\n<td>486789012<\/td>\n<td>78901234<\/td>\n<td>0765432100<\/td>\n<td>345678901200<\/td>\n<td>17<\/td>\n<td>1302<\/td>\n<td>0h1m32s<\/td>\n<\/tr>\n<tr>\n<td>11<\/td>\n<td>5.95<\/td>\n<td>28.2<\/td>\n<td>497890123<\/td>\n<td>67890123<\/td>\n<td>1543210980<\/td>\n<td>456789012300<\/td>\n<td>38<\/td>\n<td>1005<\/td>\n<td>0h1m33s<\/td>\n<\/tr>\n<tr>\n<td>12<\/td>\n<td>6.0<\/td>\n<td>28.3<\/td>\n<td>508901234<\/td>\n<td>56789012<\/td>\n<td>2109876540<\/td>\n<td>567890123400<\/td>\n<td>6<\/td>\n<td>1552<\/td>\n<td>0h1m34s<\/td>\n<\/tr>\n<tr>\n<td>13<\/td>\n<td>6.05<\/td>\n<td>28.4<\/td>\n<td>519234567<\/td>\n<td>45678901<\/td>\n<td>2678901230<\/td>\n<td>678901234500<\/td>\n<td>8<\/td>\n<td>1097<\/td>\n<td>0h1m35s<\/td>\n<\/tr>\n<tr>\n<td>14<\/td>\n<td>6.1<\/td>\n<td>28.5<\/td>\n<td>529876543<\/td>\n<td>34567890<\/td>\n<td>3109876540<\/td>\n<td>789012345600<\/td>\n<td>1173<\/td>\n<td>1689<\/td>\n<td>0h1m36s<\/td>\n<\/tr>\n<tr>\n<td>15<\/td>\n<td>6.15<\/td>\n<td>28.6<\/td>\n<td>540567890<\/td>\n<td>23456789<\/td>\n<td>3543210980<\/td>\n<td>890123456700<\/td>\n<td>27<\/td>\n<td>1400<\/td>\n<td>0h1m37s<\/td>\n<\/tr>\n<tr>\n<td>16<\/td>\n<td>6.2<\/td>\n<td>28.7<\/td>\n<td>551234567<\/td>\n<td>12345678<\/td>\n<td>3978901230<\/td>\n<td>901234567800<\/td>\n<td>583<\/td>\n<td>2019<\/td>\n<td>0h1m38s<\/td>\n<\/tr>\n<tr>\n<td>17<\/td>\n<td>6.25<\/td>\n<td>28.8<\/td>\n<td>561987654<\/td>\n<td>24678901<\/td>\n<td>4310987650<\/td>\n<td>012345678900<\/td>\n<td>19<\/td>\n<td>1365<\/td>\n<td>0h1m39s<\/td>\n<\/tr>\n<tr>\n<td>18<\/td>\n<td>6.3<\/td>\n<td>28.9<\/td>\n<td>572765432<\/td>\n<td>36789012<\/td>\n<td>4534567890<\/td>\n<td>123456789000<\/td>\n<td>20<\/td>\n<td>844<\/td>\n<td>0h1m40s<\/td>\n<\/tr>\n<tr>\n<td>19<\/td>\n<td>6.35<\/td>\n<td>29.0<\/td>\n<td>583654321<\/td>\n<td>48901234<\/td>\n<td>4660987650<\/td>\n<td>234567890100<\/td>\n<td>427<\/td>\n<td>470<\/td>\n<td>0h1m41s<\/td>\n<\/tr>\n<tr>\n<td>20<\/td>\n<td>6.4<\/td>\n<td>29.1<\/td>\n<td>594654321<\/td>\n<td>61098765<\/td>\n<td>4678901230<\/td>\n<td>345678901200<\/td>\n<td>12<\/td>\n<td>1220<\/td>\n<td>0h1m42s<\/td>\n<\/tr>\n<tr>\n<td>21<\/td>\n<td>6.45<\/td>\n<td>29.2<\/td>\n<td>605765432<\/td>\n<td>73210987<\/td>\n<td>4598765430<\/td>\n<td>456789012300<\/td>\n<td>508<\/td>\n<td>69<\/td>\n<td>0h1m43s<\/td>\n<\/tr>\n<tr>\n<td>22<\/td>\n<td>6.5<\/td>\n<td>29.3<\/td>\n<td>616987654<\/td>\n<td>85321098<\/td>\n<td>4423456780<\/td>\n<td>567890123400<\/td>\n<td>582<\/td>\n<td>93<\/td>\n<td>0h1m44s<\/td>\n<\/tr>\n<tr>\n<td>23<\/td>\n<td>6.55<\/td>\n<td>29.4<\/td>\n<td>628345678<\/td>\n<td>97432109<\/td>\n<td>4150987650<\/td>\n<td>678901234500<\/td>\n<td>14<\/td>\n<td>1018<\/td>\n<td>0h1m45s<\/td>\n<\/tr>\n<tr>\n<td>24<\/td>\n<td>6.6<\/td>\n<td>29.5<\/td>\n<td>639876543<\/td>\n<td>10954321<\/td>\n<td>3789012340<\/td>\n<td>789012345600<\/td>\n<td>12<\/td>\n<td>1748<\/td>\n<td>0h1m46s<\/td>\n<\/tr>\n<tr>\n<td>25<\/td>\n<td>6.65<\/td>\n<td>29.6<\/td>\n<td>651567890<\/td>\n<td>12165432<\/td>\n<td>3338901230<\/td>\n<td>890123456700<\/td>\n<td>2002<\/td>\n<td>1468<\/td>\n<td>0h1m47s<\/td>\n<\/tr>\n<tr>\n<td>26<\/td>\n<td>6.7<\/td>\n<td>29.7<\/td>\n<td>663456789<\/td>\n<td>13376543<\/td>\n<td>2802345670<\/td>\n<td>987654321000<\/td>\n<td>157<\/td>\n<td>1617<\/td>\n<td>0h1m48s<\/td>\n<\/tr>\n<tr>\n<td>27<\/td>\n<td>6.75<\/td>\n<td>29.8<\/td>\n<td>675567890<\/td>\n<td>14587654<\/td>\n<td>2178901230<\/td>\n<td>076543210900<\/td>\n<td>2028<\/td>\n<td>1883<\/td>\n<td>0h1m49s<\/td>\n<\/tr>\n<tr>\n<td>28<\/td>\n<td>6.8<\/td>\n<td>29.9<\/td>\n<td>687890123<\/td>\n<td>15798765<\/td>\n<td>1469012340<\/td>\n<td>156789012300<\/td>\n<td>1338<\/td>\n<td>1463<\/td>\n<td>0h1m50s<\/td>\n<\/tr>\n<tr>\n<td>29<\/td>\n<td>6.85<\/td>\n<td>30.0<\/td>\n<td>699321098<\/td>\n<td>16909876<\/td>\n<td>0668901230<\/td>\n<td>236789012300<\/td>\n<td>679<\/td>\n<td>691<\/td>\n<td>0h1m51s<\/td>\n<\/tr>\n<tr>\n<td>30<\/td>\n<td>6.9<\/td>\n<td>30.1<\/td>\n<td>711845678<\/td>\n<td>18020987<\/td>\n<td>0772345670<\/td>\n<td>316789012300<\/td>\n<td>112<\/td>\n<td>695<\/td>\n<td>0h1m52s<\/td>\n<\/tr>\n<tr>\n<td>31<\/td>\n<td>6.95<\/td>\n<td>30.2<\/td>\n<td>724456789<\/td>\n<td>19132109<\/td>\n<td>0779876540<\/td>\n<td>396789012300<\/td>\n<td>29<\/td>\n<td>371<\/td>\n<td>0h1m53s<\/td>\n<\/tr>\n<tr>\n<td>32<\/td>\n<td>7.0<\/td>\n<td>30.3<\/td>\n<td>737267890<\/td>\n<td>20243210<\/td>\n<td>0690987650<\/td>\n<td>476789012300<\/td>\n<td>38<\/td>\n<td>1005<\/td>\n<td>0h1m54s<\/td>\n<\/tr>\n<tr>\n<td>33<\/td>\n<td>7.05<\/td>\n<td>30.4<\/td>\n<td>750267890<\/td>\n<td>21354321<\/td>\n<td>0496789010<\/td>\n<td>556789012300<\/td>\n<td>6<\/td>\n<td>1552<\/td>\n<td>0h1m55s<\/td>\n<\/tr>\n<tr>\n<td>34<\/td>\n<td>7.1<\/td>\n<td>30.5<\/td>\n<td>763456789<\/td>\n<td>22465432<\/td>\n<td>0216789010<\/td>\n<td>636789012300<\/td>\n<td>8<\/td>\n<td>1097<\/td>\n<td>0h1m56s<\/td>\n<\/tr>\n<tr>\n<td>35<\/td>\n<td>7.15<\/td>\n<td>30.6<\/td>\n<td>776890123<\/td>\n<td>23576543<\/td>\n<td>0821678900<\/td>\n<td>716789012300<\/td>\n<td>1173<\/td>\n<td>1689<\/td>\n<td>0h1m57s<\/td>\n<\/tr>\n<tr>\n<td>36<\/td>\n<td>7.2<\/td>\n<td>30.7<\/td>\n<td>790567890<\/td>\n<td>24687654<\/td>\n<td>0311678900<\/td>\n<td>796789012300<\/td>\n<td>27<\/td>\n<td>1400<\/td>\n<td>0h1m58s<\/td>\n<\/tr>\n<tr>\n<td>37<\/td>\n<td>7.25<\/td>\n<td>30.8<\/td>\n<td>804456789<\/td>\n<td>25798765<\/td>\n<td>0701678900<\/td>\n<td>876789012300<\/td>\n<td>583<\/td>\n<td>2019<\/td>\n<td>0h1m59s<\/td>\n<\/tr>\n<tr>\n<td>38<\/td>\n<td>7.3<\/td>\n<td>30.9<\/td>\n<td>818567890<\/td>\n<td>26909876<\/td>\n<td>0985678900<\/td>\n<td>956789012300<\/td>\n<td>19<\/td>\n<td>1365<\/td>\n<td>0h2m0s<\/td>\n<\/tr>\n<tr>\n<td>39<\/td>\n<td>7.35<\/td>\n<td>31.0<\/td>\n<td>832901234<\/td>\n<td>28020987<\/td>\n<td>0999678900<\/td>\n<td>036789012300<\/td>\n<td>20<\/td>\n<td>844<\/td>\n<td>0h2m1s<\/td>\n<\/tr>\n<tr>\n<td>40<\/td>\n<td>7.4<\/td>\n<td>31.1<\/td>\n<td>847456789<\/td>\n<td>29132109<\/td>\n<td>0934567890<\/td>\n<td>116789012300<\/td>\n<td>427<\/td>\n<td>470<\/td>\n<td>0h2m2s<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h4>\u00fc Resultados estat\u00edsticos<\/h4>\n<ul>\n<li><strong>Tempo total aproximado<\/strong><strong> (aproximado_total <\/strong><strong>agregado<\/strong><strong>)<\/strong>: 3271,6 segundos<\/li>\n<li><strong>Tempo total m\u00e9dio aproximado<\/strong><strong> (aproximado_total <\/strong><strong>valor m\u00e9dio<\/strong><strong>)<\/strong>: 81,79 segundos<\/li>\n<\/ul>\n<h5>5. satisfa\u00e7\u00e3o do utilizador com o modelo<\/h5>\n<p>Esta an\u00e1lise utilizou v\u00e1rios utilizadores para classificar o desempenho geral dos modelos DeepSeek 70B e 7B, com cada utilizador a atribuir uma pontua\u00e7\u00e3o com base na sua pr\u00f3pria experi\u00eancia.<\/p>\n<table width=\"100%\">\n<thead>\n<tr>\n<td width=\"27%\"><strong>ID do utilizador<\/strong><\/td>\n<td width=\"37%\"><strong>70B <\/strong><strong>pontua\u00e7\u00e3o do modelo<\/strong><\/td>\n<td width=\"35%\"><strong>7B <\/strong><strong>pontua\u00e7\u00e3o do modelo<\/strong><\/td>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td width=\"27%\">1<\/td>\n<td width=\"37%\">60<\/td>\n<td width=\"35%\">70<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">2<\/td>\n<td width=\"37%\">80<\/td>\n<td width=\"35%\">60<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">3<\/td>\n<td width=\"37%\">75<\/td>\n<td width=\"35%\">40<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">4<\/td>\n<td width=\"37%\">70<\/td>\n<td width=\"35%\">40<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">5<\/td>\n<td width=\"37%\">80<\/td>\n<td width=\"35%\">60<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">6<\/td>\n<td width=\"37%\">60<\/td>\n<td width=\"35%\">60<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">7<\/td>\n<td width=\"37%\">60<\/td>\n<td width=\"35%\">70<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">8<\/td>\n<td width=\"37%\">10<\/td>\n<td width=\"35%\">30<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">9<\/td>\n<td width=\"37%\">50<\/td>\n<td width=\"35%\">70<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">10<\/td>\n<td width=\"37%\">0<\/td>\n<td width=\"35%\">60<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">11<\/td>\n<td width=\"37%\">0<\/td>\n<td width=\"35%\">50<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">12<\/td>\n<td width=\"37%\">0<\/td>\n<td width=\"35%\">40<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">13<\/td>\n<td width=\"37%\">5<\/td>\n<td width=\"35%\">10<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">14<\/td>\n<td width=\"37%\">85<\/td>\n<td width=\"35%\">60<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">15<\/td>\n<td width=\"37%\">60<\/td>\n<td width=\"35%\">50<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">16<\/td>\n<td width=\"37%\">35<\/td>\n<td width=\"35%\">20<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">17<\/td>\n<td width=\"37%\">5<\/td>\n<td width=\"35%\">60<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">18<\/td>\n<td width=\"37%\">96<\/td>\n<td width=\"35%\">80<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">19<\/td>\n<td width=\"37%\">60<\/td>\n<td width=\"35%\">60<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">20<\/td>\n<td width=\"37%\">60<\/td>\n<td width=\"35%\">20<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">21<\/td>\n<td width=\"37%\">40<\/td>\n<td width=\"35%\">20<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">22<\/td>\n<td width=\"37%\">5<\/td>\n<td width=\"35%\">5<\/td>\n<\/tr>\n<tr>\n<td width=\"27%\">Total (geral)<\/td>\n<td width=\"37%\">Pontua\u00e7\u00e3o m\u00e9dia 45,27<\/td>\n<td width=\"35%\">Pontua\u00e7\u00e3o m\u00e9dia 47,04<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h4>\u00fc Resultados estat\u00edsticos<\/h4>\n<ul>\n<li><strong>70B <\/strong><strong>Pontua\u00e7\u00e3o m\u00e9dia do modelo: 45,27<\/strong><\/li>\n<li><strong>7B <\/strong><strong>Pontua\u00e7\u00e3o m\u00e9dia do modelo: 47,05<\/strong><\/li>\n<\/ul>\n<p>Em termos de pontua\u00e7\u00f5es m\u00e9dias, n\u00e3o h\u00e1 grande diferen\u00e7a entre os dois, e a satisfa\u00e7\u00e3o geral com o desempenho do modelo 7b \u00e9 ligeiramente melhor do que a do modelo 70b, mas temos de ter em conta que o modelo 70b tem baixas classifica\u00e7\u00f5es dos utilizadores devido a uma resposta demasiado lenta, e os resultados n\u00e3o s\u00e3o suficientemente objectivos.<br \/>\nAqui est\u00e1 a sua tabela optimizada com uma formata\u00e7\u00e3o melhorada, onde tanto \"Ver mais produtos\" como \"Ver mais conte\u00fado\" est\u00e3o agora tamb\u00e9m ligados. \" est\u00e3o agora tamb\u00e9m ligados.<\/p>\n<p>&nbsp;<\/p>\n<table>\n<thead>\n<tr>\n<th>\n<h4>Para mais produtos, consultar<\/h4>\n<\/th>\n<th>\n<h4>Ver mais em<\/h4>\n<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><a href=\"https:\/\/www.myshirtai.com\/pt\/\">ShirtAI - Intelig\u00eancia penetrante<\/a><\/td>\n<td><a href=\"https:\/\/www.myshirtai.com\/pt\/archives\/4425\/\">O Grande Modelo do AIGC: inaugurando uma era de dupla revolu\u00e7\u00e3o na engenharia e na ci\u00eancia - Penetrating Intelligence<\/a><\/td>\n<\/tr>\n<tr>\n<td><a href=\"https:\/\/www.myshirtai.com\/pt\/\">1:1 Restaura\u00e7\u00e3o de Claude e GPT Site oficial - AI Cloud Native<\/a><\/td>\n<td><a href=\"https:\/\/www.bluelsqkj.com\/archives\/2876\">Aplica\u00e7\u00e3o de jogos em direto Leitor de visualiza\u00e7\u00e3o de desporto HD global (recomendado) - Blueshirt Technology<\/a><\/td>\n<\/tr>\n<tr>\n<td><a href=\"https:\/\/api.mygptmeta.com\/\">Servi\u00e7o de tr\u00e2nsito baseado na API oficial - API GPTMeta<\/a><\/td>\n<td><a href=\"https:\/\/www.zhihu.com\/question\/621055223\/answer\/3633615705\">Ajuda, algu\u00e9m pode dar algumas dicas sobre como fazer perguntas no GPT? - Conhecimento<\/a><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>&nbsp;<\/p>","protected":false},"excerpt":{"rendered":"<p>\u4e00\u3001\u7814\u7a76\u7ed3\u8bba 1.\u603b\u4f53\u7ed3\u8bba \u672c\u6b21\u7814\u7a76\u7ed3\u679c\u8868\u660e\uff0c\u5728\u672c\u5730\u76ee\u524d\u80fd\u627e\u5230\u7684\u8f83\u9ad8\u7b97\u529b\u6761\u4ef6\u4e0b\uff0c\u8fd0\u884c DeepSeek\u57fa\u7840\u7248\u6a21 [&hellip;]<\/p>","protected":false},"author":1,"featured_media":5523,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_uag_custom_page_level_css":"","site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"disabled","footer-sml-layout":"","theme-transparent-header-meta":"default","adv-header-id-meta":"","stick-header-meta":"default","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"set","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[37],"tags":[],"class_list":["post-5517","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized-zh-cn"],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v22.3 (Yoast SEO v25.2) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>\u672c\u5730DeepSeek\u6a21\u578b\u9700\u8981\u4ec0\u4e48\u914d\u7f6e\u4ee5\u53ca\u5404\u914d\u7f6e\u8dd1\u5206 - \u6e17\u900f\u667a\u80fd<\/title>\n<meta name=\"description\" content=\"\u672c\u5730\u90e8\u7f72DeepSeek\u6a21\u578b\u9700\u8981\u9ad8\u6027\u80fd\u7684\u786c\u4ef6\u914d\u7f6e\uff0c\u5305\u62ecGPU\u663e\u5b58\u3001CPU\u6838\u5fc3\u6570\u548c\u5185\u5b58\u7b49\u56e0\u7d20\u90fd\u4f1a\u5f71\u54cd\u8fd0\u884c\u6548\u7387\u3002\u4e0d\u540c\u914d\u7f6e\u4e0b\u7684\u6027\u80fd\u8dd1\u5206\uff08\u5982\u63a8\u7406\u901f\u5ea6\u3001\u541e\u5410\u91cf\uff09\u4f1a\u6709\u6240\u4e0d\u540c\uff0c\u901a\u5e38\u9ad8\u7aef\u663e\u5361\uff08\u5982A100\u30013090\u30014090\uff09\u5728\u63a8\u7406\u901f\u5ea6\u4e0a\u66f4\u5177\u4f18\u52bf\uff0c\u800cCPU\u548c\u5185\u5b58\u7684\u4f18\u5316\u4e5f\u80fd\u63d0\u5347\u6574\u4f53\u7a33\u5b9a\u6027\u3002\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.myshirtai.com\/pt\/archives\/5517\/\" \/>\n<meta property=\"og:locale\" content=\"pt_PT\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"\u672c\u5730DeepSeek\u6a21\u578b\u9700\u8981\u4ec0\u4e48\u914d\u7f6e\u4ee5\u53ca\u5404\u914d\u7f6e\u8dd1\u5206\" \/>\n<meta property=\"og:description\" content=\"\u672c\u5730\u90e8\u7f72DeepSeek\u6a21\u578b\u9700\u8981\u9ad8\u6027\u80fd\u7684\u786c\u4ef6\u914d\u7f6e\uff0c\u5305\u62ecGPU\u663e\u5b58\u3001CPU\u6838\u5fc3\u6570\u548c\u5185\u5b58\u7b49\u56e0\u7d20\u90fd\u4f1a\u5f71\u54cd\u8fd0\u884c\u6548\u7387\u3002\u4e0d\u540c\u914d\u7f6e\u4e0b\u7684\u6027\u80fd\u8dd1\u5206\uff08\u5982\u63a8\u7406\u901f\u5ea6\u3001\u541e\u5410\u91cf\uff09\u4f1a\u6709\u6240\u4e0d\u540c\uff0c\u901a\u5e38\u9ad8\u7aef\u663e\u5361\uff08\u5982A100\u30013090\u30014090\uff09\u5728\u63a8\u7406\u901f\u5ea6\u4e0a\u66f4\u5177\u4f18\u52bf\uff0c\u800cCPU\u548c\u5185\u5b58\u7684\u4f18\u5316\u4e5f\u80fd\u63d0\u5347\u6574\u4f53\u7a33\u5b9a\u6027\u3002\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.myshirtai.com\/pt\/archives\/5517\/\" \/>\n<meta property=\"og:site_name\" content=\"\u6e17\u900f\u667a\u80fd\" \/>\n<meta property=\"article:published_time\" content=\"2025-02-19T10:09:18+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2025-02-19T10:23:53+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0-1024x607.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1024\" \/>\n\t<meta property=\"og:image:height\" content=\"607\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"IvesFeng666\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Escrito por\" \/>\n\t<meta name=\"twitter:data1\" content=\"IvesFeng666\" \/>\n\t<meta name=\"twitter:label2\" content=\"Tempo estimado de leitura\" \/>\n\t<meta name=\"twitter:data2\" content=\"11 minutos\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/www.myshirtai.com\/archives\/5517#article\",\"isPartOf\":{\"@id\":\"https:\/\/www.myshirtai.com\/archives\/5517\"},\"author\":{\"name\":\"IvesFeng666\",\"@id\":\"https:\/\/www.myshirtai.com\/#\/schema\/person\/793ffae65b0212a937f22250e83b51e2\"},\"headline\":\"\u672c\u5730DeepSeek\u6a21\u578b\u9700\u8981\u4ec0\u4e48\u914d\u7f6e\u4ee5\u53ca\u5404\u914d\u7f6e\u8dd1\u5206\",\"datePublished\":\"2025-02-19T10:09:18+00:00\",\"dateModified\":\"2025-02-19T10:23:53+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/www.myshirtai.com\/archives\/5517\"},\"wordCount\":688,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\/\/www.myshirtai.com\/#organization\"},\"image\":{\"@id\":\"https:\/\/www.myshirtai.com\/archives\/5517#primaryimage\"},\"thumbnailUrl\":\"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png\",\"articleSection\":[\"Uncategorized\"],\"inLanguage\":\"pt-PT\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/www.myshirtai.com\/archives\/5517#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.myshirtai.com\/archives\/5517\",\"url\":\"https:\/\/www.myshirtai.com\/archives\/5517\",\"name\":\"\u672c\u5730DeepSeek\u6a21\u578b\u9700\u8981\u4ec0\u4e48\u914d\u7f6e\u4ee5\u53ca\u5404\u914d\u7f6e\u8dd1\u5206 - \u6e17\u900f\u667a\u80fd\",\"isPartOf\":{\"@id\":\"https:\/\/www.myshirtai.com\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/www.myshirtai.com\/archives\/5517#primaryimage\"},\"image\":{\"@id\":\"https:\/\/www.myshirtai.com\/archives\/5517#primaryimage\"},\"thumbnailUrl\":\"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png\",\"datePublished\":\"2025-02-19T10:09:18+00:00\",\"dateModified\":\"2025-02-19T10:23:53+00:00\",\"description\":\"\u672c\u5730\u90e8\u7f72DeepSeek\u6a21\u578b\u9700\u8981\u9ad8\u6027\u80fd\u7684\u786c\u4ef6\u914d\u7f6e\uff0c\u5305\u62ecGPU\u663e\u5b58\u3001CPU\u6838\u5fc3\u6570\u548c\u5185\u5b58\u7b49\u56e0\u7d20\u90fd\u4f1a\u5f71\u54cd\u8fd0\u884c\u6548\u7387\u3002\u4e0d\u540c\u914d\u7f6e\u4e0b\u7684\u6027\u80fd\u8dd1\u5206\uff08\u5982\u63a8\u7406\u901f\u5ea6\u3001\u541e\u5410\u91cf\uff09\u4f1a\u6709\u6240\u4e0d\u540c\uff0c\u901a\u5e38\u9ad8\u7aef\u663e\u5361\uff08\u5982A100\u30013090\u30014090\uff09\u5728\u63a8\u7406\u901f\u5ea6\u4e0a\u66f4\u5177\u4f18\u52bf\uff0c\u800cCPU\u548c\u5185\u5b58\u7684\u4f18\u5316\u4e5f\u80fd\u63d0\u5347\u6574\u4f53\u7a33\u5b9a\u6027\u3002\",\"breadcrumb\":{\"@id\":\"https:\/\/www.myshirtai.com\/archives\/5517#breadcrumb\"},\"inLanguage\":\"pt-PT\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.myshirtai.com\/archives\/5517\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"pt-PT\",\"@id\":\"https:\/\/www.myshirtai.com\/archives\/5517#primaryimage\",\"url\":\"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png\",\"contentUrl\":\"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png\",\"width\":4702,\"height\":2787},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.myshirtai.com\/archives\/5517#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.myshirtai.com\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"\u672c\u5730DeepSeek\u6a21\u578b\u9700\u8981\u4ec0\u4e48\u914d\u7f6e\u4ee5\u53ca\u5404\u914d\u7f6e\u8dd1\u5206\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.myshirtai.com\/#website\",\"url\":\"https:\/\/www.myshirtai.com\/\",\"name\":\"\u6e17\u900f\u667a\u80fd\",\"description\":\"ShirtAI\",\"publisher\":{\"@id\":\"https:\/\/www.myshirtai.com\/#organization\"},\"alternateName\":\"ShirtAI\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.myshirtai.com\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"pt-PT\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/www.myshirtai.com\/#organization\",\"name\":\"ShirtAI\",\"alternateName\":\"ShirtAI\",\"url\":\"https:\/\/www.myshirtai.com\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"pt-PT\",\"@id\":\"https:\/\/www.myshirtai.com\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2023\/11\/ShirtAI1279\u00d7675.png\",\"contentUrl\":\"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2023\/11\/ShirtAI1279\u00d7675.png\",\"width\":1200,\"height\":675,\"caption\":\"ShirtAI\"},\"image\":{\"@id\":\"https:\/\/www.myshirtai.com\/#\/schema\/logo\/image\/\"}},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.myshirtai.com\/#\/schema\/person\/793ffae65b0212a937f22250e83b51e2\",\"name\":\"IvesFeng666\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"pt-PT\",\"@id\":\"https:\/\/www.myshirtai.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/0e40122f3ea588c331477d2b5778ab521f0ef9275880700b47f592c999e721b7?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/0e40122f3ea588c331477d2b5778ab521f0ef9275880700b47f592c999e721b7?s=96&d=mm&r=g\",\"caption\":\"IvesFeng666\"},\"sameAs\":[\"http:\/\/www.myshirtai.com\"],\"url\":\"https:\/\/www.myshirtai.com\/pt\/archives\/author\/admin\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"\u672c\u5730DeepSeek\u6a21\u578b\u9700\u8981\u4ec0\u4e48\u914d\u7f6e\u4ee5\u53ca\u5404\u914d\u7f6e\u8dd1\u5206 - \u6e17\u900f\u667a\u80fd","description":"\u672c\u5730\u90e8\u7f72DeepSeek\u6a21\u578b\u9700\u8981\u9ad8\u6027\u80fd\u7684\u786c\u4ef6\u914d\u7f6e\uff0c\u5305\u62ecGPU\u663e\u5b58\u3001CPU\u6838\u5fc3\u6570\u548c\u5185\u5b58\u7b49\u56e0\u7d20\u90fd\u4f1a\u5f71\u54cd\u8fd0\u884c\u6548\u7387\u3002\u4e0d\u540c\u914d\u7f6e\u4e0b\u7684\u6027\u80fd\u8dd1\u5206\uff08\u5982\u63a8\u7406\u901f\u5ea6\u3001\u541e\u5410\u91cf\uff09\u4f1a\u6709\u6240\u4e0d\u540c\uff0c\u901a\u5e38\u9ad8\u7aef\u663e\u5361\uff08\u5982A100\u30013090\u30014090\uff09\u5728\u63a8\u7406\u901f\u5ea6\u4e0a\u66f4\u5177\u4f18\u52bf\uff0c\u800cCPU\u548c\u5185\u5b58\u7684\u4f18\u5316\u4e5f\u80fd\u63d0\u5347\u6574\u4f53\u7a33\u5b9a\u6027\u3002","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.myshirtai.com\/pt\/archives\/5517\/","og_locale":"pt_PT","og_type":"article","og_title":"\u672c\u5730DeepSeek\u6a21\u578b\u9700\u8981\u4ec0\u4e48\u914d\u7f6e\u4ee5\u53ca\u5404\u914d\u7f6e\u8dd1\u5206","og_description":"\u672c\u5730\u90e8\u7f72DeepSeek\u6a21\u578b\u9700\u8981\u9ad8\u6027\u80fd\u7684\u786c\u4ef6\u914d\u7f6e\uff0c\u5305\u62ecGPU\u663e\u5b58\u3001CPU\u6838\u5fc3\u6570\u548c\u5185\u5b58\u7b49\u56e0\u7d20\u90fd\u4f1a\u5f71\u54cd\u8fd0\u884c\u6548\u7387\u3002\u4e0d\u540c\u914d\u7f6e\u4e0b\u7684\u6027\u80fd\u8dd1\u5206\uff08\u5982\u63a8\u7406\u901f\u5ea6\u3001\u541e\u5410\u91cf\uff09\u4f1a\u6709\u6240\u4e0d\u540c\uff0c\u901a\u5e38\u9ad8\u7aef\u663e\u5361\uff08\u5982A100\u30013090\u30014090\uff09\u5728\u63a8\u7406\u901f\u5ea6\u4e0a\u66f4\u5177\u4f18\u52bf\uff0c\u800cCPU\u548c\u5185\u5b58\u7684\u4f18\u5316\u4e5f\u80fd\u63d0\u5347\u6574\u4f53\u7a33\u5b9a\u6027\u3002","og_url":"https:\/\/www.myshirtai.com\/pt\/archives\/5517\/","og_site_name":"\u6e17\u900f\u667a\u80fd","article_published_time":"2025-02-19T10:09:18+00:00","article_modified_time":"2025-02-19T10:23:53+00:00","og_image":[{"width":1024,"height":607,"url":"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0-1024x607.png","type":"image\/png"}],"author":"IvesFeng666","twitter_card":"summary_large_image","twitter_misc":{"Escrito por":"IvesFeng666","Tempo estimado de leitura":"11 minutos"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.myshirtai.com\/archives\/5517#article","isPartOf":{"@id":"https:\/\/www.myshirtai.com\/archives\/5517"},"author":{"name":"IvesFeng666","@id":"https:\/\/www.myshirtai.com\/#\/schema\/person\/793ffae65b0212a937f22250e83b51e2"},"headline":"\u672c\u5730DeepSeek\u6a21\u578b\u9700\u8981\u4ec0\u4e48\u914d\u7f6e\u4ee5\u53ca\u5404\u914d\u7f6e\u8dd1\u5206","datePublished":"2025-02-19T10:09:18+00:00","dateModified":"2025-02-19T10:23:53+00:00","mainEntityOfPage":{"@id":"https:\/\/www.myshirtai.com\/archives\/5517"},"wordCount":688,"commentCount":0,"publisher":{"@id":"https:\/\/www.myshirtai.com\/#organization"},"image":{"@id":"https:\/\/www.myshirtai.com\/archives\/5517#primaryimage"},"thumbnailUrl":"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png","articleSection":["Uncategorized"],"inLanguage":"pt-PT","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/www.myshirtai.com\/archives\/5517#respond"]}]},{"@type":"WebPage","@id":"https:\/\/www.myshirtai.com\/archives\/5517","url":"https:\/\/www.myshirtai.com\/archives\/5517","name":"\u672c\u5730DeepSeek\u6a21\u578b\u9700\u8981\u4ec0\u4e48\u914d\u7f6e\u4ee5\u53ca\u5404\u914d\u7f6e\u8dd1\u5206 - \u6e17\u900f\u667a\u80fd","isPartOf":{"@id":"https:\/\/www.myshirtai.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.myshirtai.com\/archives\/5517#primaryimage"},"image":{"@id":"https:\/\/www.myshirtai.com\/archives\/5517#primaryimage"},"thumbnailUrl":"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png","datePublished":"2025-02-19T10:09:18+00:00","dateModified":"2025-02-19T10:23:53+00:00","description":"\u672c\u5730\u90e8\u7f72DeepSeek\u6a21\u578b\u9700\u8981\u9ad8\u6027\u80fd\u7684\u786c\u4ef6\u914d\u7f6e\uff0c\u5305\u62ecGPU\u663e\u5b58\u3001CPU\u6838\u5fc3\u6570\u548c\u5185\u5b58\u7b49\u56e0\u7d20\u90fd\u4f1a\u5f71\u54cd\u8fd0\u884c\u6548\u7387\u3002\u4e0d\u540c\u914d\u7f6e\u4e0b\u7684\u6027\u80fd\u8dd1\u5206\uff08\u5982\u63a8\u7406\u901f\u5ea6\u3001\u541e\u5410\u91cf\uff09\u4f1a\u6709\u6240\u4e0d\u540c\uff0c\u901a\u5e38\u9ad8\u7aef\u663e\u5361\uff08\u5982A100\u30013090\u30014090\uff09\u5728\u63a8\u7406\u901f\u5ea6\u4e0a\u66f4\u5177\u4f18\u52bf\uff0c\u800cCPU\u548c\u5185\u5b58\u7684\u4f18\u5316\u4e5f\u80fd\u63d0\u5347\u6574\u4f53\u7a33\u5b9a\u6027\u3002","breadcrumb":{"@id":"https:\/\/www.myshirtai.com\/archives\/5517#breadcrumb"},"inLanguage":"pt-PT","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.myshirtai.com\/archives\/5517"]}]},{"@type":"ImageObject","inLanguage":"pt-PT","@id":"https:\/\/www.myshirtai.com\/archives\/5517#primaryimage","url":"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png","contentUrl":"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png","width":4702,"height":2787},{"@type":"BreadcrumbList","@id":"https:\/\/www.myshirtai.com\/archives\/5517#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.myshirtai.com\/"},{"@type":"ListItem","position":2,"name":"\u672c\u5730DeepSeek\u6a21\u578b\u9700\u8981\u4ec0\u4e48\u914d\u7f6e\u4ee5\u53ca\u5404\u914d\u7f6e\u8dd1\u5206"}]},{"@type":"WebSite","@id":"https:\/\/www.myshirtai.com\/#website","url":"https:\/\/www.myshirtai.com\/","name":"\u6e17\u900f\u667a\u80fd","description":"ShirtAI","publisher":{"@id":"https:\/\/www.myshirtai.com\/#organization"},"alternateName":"ShirtAI","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.myshirtai.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"pt-PT"},{"@type":"Organization","@id":"https:\/\/www.myshirtai.com\/#organization","name":"ShirtAI","alternateName":"ShirtAI","url":"https:\/\/www.myshirtai.com\/","logo":{"@type":"ImageObject","inLanguage":"pt-PT","@id":"https:\/\/www.myshirtai.com\/#\/schema\/logo\/image\/","url":"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2023\/11\/ShirtAI1279\u00d7675.png","contentUrl":"https:\/\/www.myshirtai.com\/wp-content\/uploads\/2023\/11\/ShirtAI1279\u00d7675.png","width":1200,"height":675,"caption":"ShirtAI"},"image":{"@id":"https:\/\/www.myshirtai.com\/#\/schema\/logo\/image\/"}},{"@type":"Person","@id":"https:\/\/www.myshirtai.com\/#\/schema\/person\/793ffae65b0212a937f22250e83b51e2","name":"IvesFeng666","image":{"@type":"ImageObject","inLanguage":"pt-PT","@id":"https:\/\/www.myshirtai.com\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/0e40122f3ea588c331477d2b5778ab521f0ef9275880700b47f592c999e721b7?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/0e40122f3ea588c331477d2b5778ab521f0ef9275880700b47f592c999e721b7?s=96&d=mm&r=g","caption":"IvesFeng666"},"sameAs":["http:\/\/www.myshirtai.com"],"url":"https:\/\/www.myshirtai.com\/pt\/archives\/author\/admin"}]}},"uagb_featured_image_src":{"full":["https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png",4702,2787,false],"thumbnail":["https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0-150x150.png",150,150,true],"medium":["https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0-300x178.png",300,178,true],"medium_large":["https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0-768x455.png",768,455,true],"large":["https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0-1024x607.png",1024,607,true],"1536x1536":["https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png",1536,910,false],"2048x2048":["https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png",2048,1214,false],"trp-custom-language-flag":["https:\/\/www.myshirtai.com\/wp-content\/uploads\/2025\/02\/e44d096e-fa46-4cae-b2f2-53991e8c8da0.png",18,12,false]},"uagb_author_info":{"display_name":"IvesFeng666","author_link":"https:\/\/www.myshirtai.com\/pt\/archives\/author\/admin"},"uagb_comment_info":0,"uagb_excerpt":"\u4e00\u3001\u7814\u7a76\u7ed3\u8bba 1.\u603b\u4f53\u7ed3\u8bba \u672c\u6b21\u7814\u7a76\u7ed3\u679c\u8868\u660e\uff0c\u5728\u672c\u5730\u76ee\u524d\u80fd\u627e\u5230\u7684\u8f83\u9ad8\u7b97\u529b\u6761\u4ef6\u4e0b\uff0c\u8fd0\u884c DeepSeek\u57fa\u7840\u7248\u6a21&hellip;","_links":{"self":[{"href":"https:\/\/www.myshirtai.com\/pt\/wp-json\/wp\/v2\/posts\/5517","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.myshirtai.com\/pt\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.myshirtai.com\/pt\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.myshirtai.com\/pt\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.myshirtai.com\/pt\/wp-json\/wp\/v2\/comments?post=5517"}],"version-history":[{"count":0,"href":"https:\/\/www.myshirtai.com\/pt\/wp-json\/wp\/v2\/posts\/5517\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.myshirtai.com\/pt\/wp-json\/wp\/v2\/media\/5523"}],"wp:attachment":[{"href":"https:\/\/www.myshirtai.com\/pt\/wp-json\/wp\/v2\/media?parent=5517"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.myshirtai.com\/pt\/wp-json\/wp\/v2\/categories?post=5517"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.myshirtai.com\/pt\/wp-json\/wp\/v2\/tags?post=5517"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}