blocks|key|4243451|text|您共享的文档包含有关Google+AI+Platform+Prediction的大量信息。总之，它是GCP中的一个托管服务，您不需要管理基础设施。您只需部署您的模型，一个新的REST端点将可供您开始通过SDK或API发送预测。|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|4243452|Docker+|4243453|支持多个框架：|unordered-list-item|4243454|TensorFlow|4243455|scikit-learn|4243456|XGBoost|4243457|Pytorch|4243458|Custom+Docker+containers+(soon)|4243459|4243460|4243461|​|4243462|4243463|支持|4243464|GPU型号versions|4243465|Online和Batch+prediction|4243466|Logging以及Monitoring|4243467|Multiple区域|4243468|REST|4243469|4243470|4243471|4243472|4243473|KFServing您需要管理自己的K8s/KubeFlow基础设施。Kubeflow支持两个允许多框架模型服务的模型服务系统:+KFServing和Seldon+Core。检查What+If+tool和型号logging.|ordered-list-item|4243474|4243475|entityMap|0|LINK|mutability|MUTABLE|url|https://cloud.google.com/ai-platform/prediction/docs/overview#how_it_works|1|https://cloud.google.com/ai-platform/prediction/docs/deploying-models#adjust_prediction_outputs_for_the_what-if_tool|2|https://cloud.google.com/ai-platform/prediction/docs/overview#prediction_logging^0|A|T|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|2G|C|1|2V|7|2|0|0^^$0|@$1|2|3|4|5|6|7|1T|8|@]|9|@$A|1U|B|1V|1|1W]]|C|$]]|$1|D|3|E|5|6|7|1X|8|@]|9|@]|C|$]]|$1|F|3|G|5|H|7|1Y|8|@]|9|@]|C|$]]|$1|I|3|J|5|H|7|1Z|8|@]|9|@]|C|$]]|$1|K|3|L|5|H|7|20|8|@]|9|@]|C|$]]|$1|M|3|N|5|H|7|21|8|@]|9|@]|C|$]]|$1|O|3|P|5|H|7|22|8|@]|9|@]|C|$]]|$1|Q|3|R|5|H|7|23|8|@]|9|@]|C|$]]|$1|S|3|-4|5|6|7|24|8|@]|9|@]|C|$]]|$1|T|3|-4|5|6|7|25|8|@]|9|@]|C|$]]|$1|U|3|V|5|6|7|26|8|@]|9|@]|C|$]]|$1|W|3|-4|5|6|7|27|8|@]|9|@]|C|$]]|$1|X|3|Y|5|H|7|28|8|@]|9|@]|C|$]]|$1|Z|3|10|5|6|7|29|8|@]|9|@]|C|$]]|$1|11|3|12|5|H|7|2A|8|@]|9|@]|C|$]]|$1|13|3|14|5|H|7|2B|8|@]|9|@]|C|$]]|$1|15|3|16|5|H|7|2C|8|@]|9|@]|C|$]]|$1|17|3|18|5|H|7|2D|8|@]|9|@]|C|$]]|$1|19|3|-4|5|H|7|2E|8|@]|9|@]|C|$]]|$1|1A|3|-4|5|6|7|2F|8|@]|9|@]|C|$]]|$1|1B|3|-4|5|6|7|2G|8|@]|9|@]|C|$]]|$1|1C|3|-4|5|6|7|2H|8|@]|9|@]|C|$]]|$1|1D|3|1E|5|1F|7|2I|8|@]|9|@$A|2J|B|2K|1|2L]|$A|2M|B|2N|1|2O]]|C|$]]|$1|1G|3|-4|5|6|7|2P|8|@]|9|@]|C|$]]|$1|1H|3|-4|5|6|7|2Q|8|@]|9|@]|C|$]]]|1I|$1J|$5|1K|1L|1M|C|$1N|1O]]|1P|$5|1K|1L|1M|C|$1N|1Q]]|1R|$5|1K|1L|1M|C|$1N|1S]]]]

The document you shared contains extensive information about <a href="https://cloud.google.com/ai-platform/prediction/docs/overview#how_it_works" rel="nofollow noreferrer">Google AI Platform Prediction</a>. In summary, it is a hosted service in GCP where you don't need to manage the infrastructure. You just deploy your model and a new REST endpoint will be available for you to start sending predictions via SDK or API.

<ul>
<li>Supports multiple frameworks: 

<ul>
<li>TensorFlow</li>
<li>scikit-learn</li>
<li>XGBoost</li>
<li>Pytorch</li>
<li>Custom Docker containers (soon)</li>
</ul></li>
<li>Support GPUs</li>
<li>Model versions</li>
<li>Online and Batch prediction</li>
<li>Logging and Monitoring</li>
<li>Multiple Regions</li>
<li>REST API</li>
</ul>

Answer to your questions:

<ol>
<li>KFServing you need to manage your own K8s/KubeFlow infrastructure.
Kubeflow supports two model serving systems that allow multi-framework model serving: KFServing and Seldon Core. 
AI Platform Service you don't manage the infrastructure, nor need K8s/KF, you simply deploy your models and GCP takes care of the infra.</li>
<li><code>gcloud ai-platform versions create</code> will deploy a VM(s) in Google Cloud where based on the settings (Runtime) and Framework all the dependencies will be installed automatically, also all you need to load your model will be installed so you can have access to a REST API.</li>
<li>Canary can be implemented used with different Models and versions, it may depend on routing your predictions. Check the <a href="https://cloud.google.com/ai-platform/prediction/docs/deploying-models#adjust_prediction_outputs_for_the_what-if_tool" rel="nofollow noreferrer">What If tool</a> and Model <a href="https://cloud.google.com/ai-platform/prediction/docs/overview#prediction_logging" rel="nofollow noreferrer">logging</a>.</li>
</ol>

blocks|key|4243496|text|type|unstyled|depth|inlineStyleRanges|entityRanges|data|4243497|Google+AI+Platform可用于管理ML工作流中的以下阶段：|ordered-list-item|4243498|4243499|在数据上使用-Train+ML模型：|offset|length|style|BOLD|4243500|训练模型评估模型精度调整超参数|4243501|-Deploy你的训练模型。|4243502|对模型的-Send预测请求：|4243503|在线预测批量预测(仅适用于TensorFlow+)|4243504|-Monitor在持续的基础上进行预测。|4243505|-Manage您的模型和模型版本。|4243506|4243507|KFServing支持Kubernetes上的无服务器推理，并为通用机器学习(ML)框架提供高性能、高抽象的接口，如TensorFlow、XGBoost、scikit+learn、PyTorch和ONNX，以解决生产模型服务用例。|4243508|4243509|entityMap^0|0|0|0|6|C|0|0|0|E|0|4|A|0|0|0|K|0|0|H|0|0|0|0^^$0|@$1|2|3|-4|4|5|6|12|7|@]|8|@]|9|$]]|$1|A|3|B|4|C|6|13|7|@]|8|@]|9|$]]|$1|D|3|-4|4|5|6|14|7|@]|8|@]|9|$]]|$1|E|3|F|4|5|6|15|7|@$G|16|H|17|I|J]]|8|@]|9|$]]|$1|K|3|L|4|5|6|18|7|@]|8|@]|9|$]]|$1|M|3|N|4|5|6|19|7|@$G|1A|H|1B|I|J]]|8|@]|9|$]]|$1|O|3|P|4|5|6|1C|7|@$G|1D|H|1E|I|J]]|8|@]|9|$]]|$1|Q|3|R|4|5|6|1F|7|@]|8|@]|9|$]]|$1|S|3|T|4|5|6|1G|7|@$G|1H|H|1I|I|J]]|8|@]|9|$]]|$1|U|3|V|4|5|6|1J|7|@$G|1K|H|1L|I|J]]|8|@]|9|$]]|$1|W|3|-4|4|5|6|1M|7|@]|8|@]|9|$]]|$1|X|3|Y|4|C|6|1N|7|@]|8|@]|9|$]]|$1|Z|3|-4|4|5|6|1O|7|@]|8|@]|9|$]]|$1|10|3|-4|4|5|6|1P|7|@]|8|@]|9|$]]]|11|$]]

<ol>
<li>Google AI Platform can be used to manage the following stages in the ML workflow:</li>
</ol>
-Train an ML model on your data:
Train model
Evaluate model accuracy
Tune hyperparameters
-Deploy your trained model.
-Send prediction requests to your model:
Online prediction
Batch prediction (for TensorFlow only)
-Monitor the predictions on an ongoing basis.
-Manage your models and model versions.
<ol start="2">
<li>KFServing enables serverless inferencing on Kubernetes and provides performant, high abstraction interfaces for common machine learning (ML) frameworks like TensorFlow, XGBoost, scikit-learn, PyTorch, and ONNX to solve production model serving use cases.</li>
</ol>

We are going to do model serving infrastructure. I am comparing <a href="https://cloud.google.com/ai-platform/prediction/docs/overview" rel="nofollow noreferrer">Google AI Prediction</a> and <a href="https://www.kubeflow.org/docs/components/serving/kfserving/" rel="nofollow noreferrer">kfserving</a>. But I cannot find enough documents about the features of google ai serving and how it is implemented.

It seems that <code>gcloud ai-platform versions create</code> can create model version resource and start serving, which is the only point I can find.

I have three questions:

1, what is relationship between google ai serving and kfserving?

2, how <code>gcloud ai-platform versions create</code> works?

3, as for the features of google ai serving, do google ai serving provide all feature such as canary rollout, explainers, monitoring, etc listed in <a href="https://www.kubeflow.org/docs/components/serving/overview/" rel="nofollow noreferrer">https://www.kubeflow.org/docs/components/serving/overview/</a>?

Thanks

google AI model serving vs kfsering

翻译质量差，导致语言生硬或混乱。

没有提供实际的解决方法或示例。

解答不清晰，无法理解或解决问题。

页面排版不美观，阅读体验差。

文章

问答

视频

学习中心

腾讯云实验室

直播

竞赛

腾讯云代码分析专区

腾讯iOA零信任安全管理系统专区

腾讯云架构师技术同盟交流圈

腾讯云数据库专区

腾讯云顾问专区

腾讯云原生专区

腾讯混元专区

腾讯云TCE专区

腾讯云Lighthouse专区

腾讯云HAI专区

腾讯云Edgeone专区

腾讯云存储专区

腾讯云智能专区

腾讯轻联专区 

腾讯云开发专区

TAPD专区

腾讯轻量云游戏服专区

腾讯云最具价值专家

腾讯云架构师技术同盟

腾讯云创作之星

腾讯云开发者先锋

腾讯云代码助手

云原生构建

TAPD 敏捷项目管理

Cloud Studio

SDK中心

API中心

命令行工具

涵盖代码开发、场景应用、自动测试全流程，助你从零构建专属AI助手

一站式MCP教程库，解锁AI应用新玩法

我们要做的是模型服务基础设施。我正在比较和。但是我找不到足够的文档来描述google ai服务的特性以及它是如何实现的。似乎gcloud ai-platform versions create可以创建模型版本资源并开始服务，这是我唯一能找到的点。我有三个问题：谷歌人工智能服务和kfserving之间的关系是什么？2，gcloud ai-platform versions create是如何工作的？

问google AI模型服务与kfsering
EN

回答 2

Stack Overflow用户

Stack Overflow用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问google AI模型服务与kfseringEN

回答 2

Stack Overflow用户

Stack Overflow用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问google AI模型服务与kfsering
EN