yexinru/doc-exports
1
0
Fork 0
forked from docs/doc-exports
Code Pull Requests Activity
Files
d12c961b0821ec1ae58155902c2a87a5ade0a358
doc-exports/docs/modelarts/umn/CLASS.TXT.json
Lai, Weijian d12c961b08 ModelArts UMN 22.3.0 version-20240311 
Reviewed-by: Pruthi, Vineet <vineet.pruthi@t-systems.com>
Co-authored-by: Lai, Weijian <laiweijian4@huawei.com>
Co-committed-by: Lai, Weijian <laiweijian4@huawei.com>
2024-04-05 09:33:47 +00:00

2972 lines
122 KiB
JSON
Raw Blame History

[
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Service Overview",
"uri":"modelarts_01_0000.html",
"doc_type":"usermanual",
"p_code":"",
"code":"1"
},
{
"desc":"ModelArts is a one-stop AI development platform geared toward developers and data scientists of all skill levels. It enables you to rapidly build, train, and deploy model",
"product_code":"modelarts",
"title":"What Is ModelArts?",
"uri":"modelarts_01_0001.html",
"doc_type":"usermanual",
"p_code":"1",
"code":"2"
},
{
"desc":"AI engineers face challenges in the installation and configuration of various AI tools, data preparation, and model training. To address these challenges, the one-stop AI",
"product_code":"modelarts",
"title":"Functions",
"uri":"modelarts_01_0003.html",
"doc_type":"usermanual",
"p_code":"1",
"code":"3"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Basic Knowledge",
"uri":"modelarts_01_0009.html",
"doc_type":"usermanual",
"p_code":"1",
"code":"4"
},
{
"desc":"Artificial intelligence (AI) is a technology capable of simulating human cognition through machines. The core capability of AI is to make a judgment or prediction based o",
"product_code":"modelarts",
"title":"Introduction to the AI Development Lifecycle",
"uri":"modelarts_01_0010.html",
"doc_type":"usermanual",
"p_code":"4",
"code":"5"
},
{
"desc":"Machine learning is classified into supervised, unsupervised, and reinforcement learning.Supervised learning uses labeled samples to adjust the parameters of classifiers ",
"product_code":"modelarts",
"title":"Basic Concepts of AI Development",
"uri":"modelarts_01_0011.html",
"doc_type":"usermanual",
"p_code":"4",
"code":"6"
},
{
"desc":"ExeML is the process of automating model design, parameter tuning, and model training, model compression, and model deployment with the labeled data. The process is code-",
"product_code":"modelarts",
"title":"Common Concepts of ModelArts",
"uri":"modelarts_01_0005.html",
"doc_type":"usermanual",
"p_code":"4",
"code":"7"
},
{
"desc":"During AI development, massive volumes of data need to be processed, and data preparation and labeling usually take more than half of the development time. ModelArts data",
"product_code":"modelarts",
"title":"Data Management",
"uri":"modelarts_01_0012.html",
"doc_type":"usermanual",
"p_code":"4",
"code":"8"
},
{
"desc":"It is challenging to set up a development environment, select an AI algorithm framework and algorithm, debug code, install software, and accelerate hardware. To help user",
"product_code":"modelarts",
"title":"DevEnviron (Old Version)",
"uri":"modelarts_01_0013-.html",
"doc_type":"usermanual",
"p_code":"4",
"code":"9"
},
{
"desc":"This document describes the DevEnviron notebook functions of the new version.Software development is a process of reducing developer costs and improving development exper",
"product_code":"modelarts",
"title":"Introduction to Development Tools",
"uri":"modelarts_01_0013.html",
"doc_type":"usermanual",
"p_code":"4",
"code":"10"
},
{
"desc":"In addition to data and algorithms, developers spend a lot of time configuring model training parameters. Model training parameters determine the model's precision and co",
"product_code":"modelarts",
"title":"Model Training",
"uri":"modelarts_01_0014.html",
"doc_type":"usermanual",
"p_code":"4",
"code":"11"
},
{
"desc":"Generally, AI model deployment and large-scale implementation are complex.The real-time inference service features high concurrency, low latency, and elastic scaling, and",
"product_code":"modelarts",
"title":"Model Deployment",
"uri":"modelarts_01_0015.html",
"doc_type":"usermanual",
"p_code":"4",
"code":"12"
},
{
"desc":"To implement AI in various industries, AI model development must be simplified. Currently, only a few algorithm engineers and researchers are capable of AI development an",
"product_code":"modelarts",
"title":"ExeML",
"uri":"modelarts_01_0020.html",
"doc_type":"usermanual",
"p_code":"4",
"code":"13"
},
{
"desc":"ModelArts uses Object Storage Service (OBS) to securely and reliably store data and models at low costs. For more details, see Object Storage Service Console Operation Gu",
"product_code":"modelarts",
"title":"Related Services",
"uri":"modelarts_01_0006.html",
"doc_type":"usermanual",
"p_code":"1",
"code":"14"
},
{
"desc":"You can access ModelArts through the web-based management console or by using HTTPS-based application programming interfaces (APIs).Using the Management ConsoleModelArts ",
"product_code":"modelarts",
"title":"How Do I Access ModelArts?",
"uri":"modelarts_01_0007.html",
"doc_type":"usermanual",
"p_code":"1",
"code":"15"
},
{
"desc":"ModelArts is a one-stop AI development platform geared toward developers and data scientists of all skill levels. It enables you to rapidly build, train, and deploy model",
"product_code":"modelarts",
"title":"Billing",
"uri":"modelarts_01_0021.html",
"doc_type":"usermanual",
"p_code":"1",
"code":"16"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Preparations",
"uri":"modelarts_08_0000.html",
"doc_type":"usermanual",
"p_code":"",
"code":"17"
},
{
"desc":"Certain ModelArts functions require access to Object Storage Service (OBS), Software Repository for Container (SWR), and Intelligent EdgeFabric (IEF). Before using ModelA",
"product_code":"modelarts",
"title":"Configuring Access Authorization (Global Configuration)",
"uri":"modelarts_08_0007.html",
"doc_type":"usermanual",
"p_code":"17",
"code":"18"
},
{
"desc":"ModelArts uses OBS to store data, and backs up and takes snapshots for models, achieving secure, reliable storage at low costs. Before using ModelArts, create an OBS buck",
"product_code":"modelarts",
"title":"Creating an OBS Bucket",
"uri":"modelarts_08_0003.html",
"doc_type":"usermanual",
"p_code":"17",
"code":"19"
},
{
"desc":"If the domain name of a region can be resolved through the public network, skip in this section. If the domain name of a region cannot be resolved through the public netw",
"product_code":"modelarts",
"title":"(Optional) Configuring Mapping Between Domain Names and IP Addresses",
"uri":"modelarts_08_0008.html",
"doc_type":"usermanual",
"p_code":"17",
"code":"20"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"ExeML",
"uri":"modelarts_21_0000.html",
"doc_type":"usermanual",
"p_code":"",
"code":"21"
},
{
"desc":"ModelArts ExeML is a customized code-free model development tool that helps you start codeless AI application development with high flexibility. ExeML automates model des",
"product_code":"modelarts",
"title":"Introduction to ExeML",
"uri":"modelarts_21_0001.html",
"doc_type":"usermanual",
"p_code":"21",
"code":"22"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Image Classification",
"uri":"modelarts_21_0002.html",
"doc_type":"usermanual",
"p_code":"21",
"code":"23"
},
{
"desc":"Before using ModelArts ExeML to build a model, upload data to an OBS bucket.This operation uses the OBS console to upload data.Perform the following operations to import ",
"product_code":"modelarts",
"title":"Preparing Data",
"uri":"modelarts_21_0003.html",
"doc_type":"usermanual",
"p_code":"23",
"code":"24"
},
{
"desc":"ModelArts ExeML supports image classification and object detection projects. You can create any of them based on your needs. Perform the following operations to create an",
"product_code":"modelarts",
"title":"Creating a Project",
"uri":"modelarts_21_0004.html",
"doc_type":"usermanual",
"p_code":"23",
"code":"25"
},
{
"desc":"Model training requires a large number of labeled images. Therefore, before model training, add labels to the images that are not labeled. ModelArts allows you to add lab",
"product_code":"modelarts",
"title":"Labeling Data",
"uri":"modelarts_21_0005.html",
"doc_type":"usermanual",
"p_code":"23",
"code":"26"
},
{
"desc":"After labeling the images, you can train a model. You can perform model training to obtain the required image classification model. Training images must be classified int",
"product_code":"modelarts",
"title":"Training a Model",
"uri":"modelarts_21_0006.html",
"doc_type":"usermanual",
"p_code":"23",
"code":"27"
},
{
"desc":"You can deploy a model as a real-time service that provides a real-time test UI and monitoring capabilities. After model training is complete, you can deploy a version wi",
"product_code":"modelarts",
"title":"Deploying a Model as a Service",
"uri":"modelarts_21_0007.html",
"doc_type":"usermanual",
"p_code":"23",
"code":"28"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Object Detection",
"uri":"modelarts_21_0008.html",
"doc_type":"usermanual",
"p_code":"21",
"code":"29"
},
{
"desc":"Before using ModelArts ExeML to build a model, upload data to an OBS bucket.This operation uses the OBS console to upload data.Perform the following operations to import ",
"product_code":"modelarts",
"title":"Preparing Data",
"uri":"modelarts_21_0009.html",
"doc_type":"usermanual",
"p_code":"29",
"code":"30"
},
{
"desc":"ModelArts ExeML supports image classification and object detection projects. You can create any of them based on your needs. Perform the following operations to create an",
"product_code":"modelarts",
"title":"Creating a Project",
"uri":"modelarts_21_0010.html",
"doc_type":"usermanual",
"p_code":"29",
"code":"31"
},
{
"desc":"Before data labeling, consider how to design labels. The labels must correspond to the distinct characteristics of the detected images and are easy to identify (the detec",
"product_code":"modelarts",
"title":"Labeling Data",
"uri":"modelarts_21_0011.html",
"doc_type":"usermanual",
"p_code":"29",
"code":"32"
},
{
"desc":"After labeling the images, perform auto training to obtain an appropriate model version.On the ExeML page, click the name of the project that is successfully created. The",
"product_code":"modelarts",
"title":"Training a Model",
"uri":"modelarts_21_0012.html",
"doc_type":"usermanual",
"p_code":"29",
"code":"33"
},
{
"desc":"You can deploy a model as a real-time service that provides a real-time test UI and monitoring capabilities. After the model is trained, you can deploy a Completed versio",
"product_code":"modelarts",
"title":"Deploying a Model as a Service",
"uri":"modelarts_21_0013.html",
"doc_type":"usermanual",
"p_code":"29",
"code":"34"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Predictive Analytics",
"uri":"modelarts_21_0014.html",
"doc_type":"usermanual",
"p_code":"21",
"code":"35"
},
{
"desc":"Before using ModelArts to build a predictive analytics model, upload data to OBS.This operation uses the OBS client to upload data. For more information about how to crea",
"product_code":"modelarts",
"title":"Preparing Data",
"uri":"modelarts_21_0015.html",
"doc_type":"usermanual",
"p_code":"35",
"code":"36"
},
{
"desc":"ModelArts ExeML supports image classification, and object detection projects. You can create any of them based on your needs. Perform the following operations to create a",
"product_code":"modelarts",
"title":"Creating a Project",
"uri":"modelarts_21_0016.html",
"doc_type":"usermanual",
"p_code":"35",
"code":"37"
},
{
"desc":"After creating a predictive analytics project, select a label column and its data type. On the Label Data tab page, you can preview data and select the label column and i",
"product_code":"modelarts",
"title":"Selecting a Label Column",
"uri":"modelarts_21_0017.html",
"doc_type":"usermanual",
"p_code":"35",
"code":"38"
},
{
"desc":"After the data is labeled, train a model for predictive analytics. You can publish the model as a real-time inference service.On the ExeML page, click the name of the pro",
"product_code":"modelarts",
"title":"Training a Model",
"uri":"modelarts_21_0018.html",
"doc_type":"usermanual",
"p_code":"35",
"code":"39"
},
{
"desc":"You can deploy a model as a real-time service that provides a real-time test UI and monitoring capabilities. After the model is trained, you can deploy a Successful versi",
"product_code":"modelarts",
"title":"Deploying a Model as a Service",
"uri":"modelarts_21_0019.html",
"doc_type":"usermanual",
"p_code":"35",
"code":"40"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Tips",
"uri":"modelarts_21_0030.html",
"doc_type":"usermanual",
"p_code":"21",
"code":"41"
},
{
"desc":"When creating a project, select a training data path. This section describes how to quickly create an OBS bucket and folder when you select the training data path.On the ",
"product_code":"modelarts",
"title":"How Do I Quickly Create an OBS Bucket and a Folder When Creating a Project?",
"uri":"modelarts_21_0031.html",
"doc_type":"usermanual",
"p_code":"41",
"code":"42"
},
{
"desc":"To add data for an existing project, perform the following operations. The operations described in this section apply only to object detection and image classification pr",
"product_code":"modelarts",
"title":"How Do I View the Added Data in an ExeML Project?",
"uri":"modelarts_21_0032.html",
"doc_type":"usermanual",
"p_code":"41",
"code":"43"
},
{
"desc":"Each round of training generates a training version in an ExeML project. If a training result is unsatisfactory (for example, if the precision is not good enough), you ca",
"product_code":"modelarts",
"title":"How Do I Perform Incremental Training in an ExeML Project?",
"uri":"modelarts_21_0033.html",
"doc_type":"usermanual",
"p_code":"41",
"code":"44"
},
{
"desc":"For an ExeML project, after the model training is complete, the generated model is automatically displayed on the AI Application Management > AI Applications page. The mo",
"product_code":"modelarts",
"title":"Where Are Models Generated by ExeML Stored? What Other Operations Are Supported?",
"uri":"modelarts_21_0034.html",
"doc_type":"usermanual",
"p_code":"41",
"code":"45"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Data Management",
"uri":"modelarts_23_0002.html",
"doc_type":"usermanual",
"p_code":"",
"code":"46"
},
{
"desc":"In ModelArts, you can import and label data on the Data Management page to prepare for model building. ModelArts uses datasets as the basis for model development or train",
"product_code":"modelarts",
"title":"Introduction to Data Management",
"uri":"modelarts_23_0003.html",
"doc_type":"usermanual",
"p_code":"46",
"code":"47"
},
{
"desc":"To manage data using ModelArts, you need to create a dataset first. Then you can perform operations on the dataset, such as labeling data, importing data, and publishing ",
"product_code":"modelarts",
"title":"Creating a Dataset",
"uri":"modelarts_23_0004.html",
"doc_type":"usermanual",
"p_code":"46",
"code":"48"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Labeling Data",
"uri":"modelarts_23_0010.html",
"doc_type":"usermanual",
"p_code":"46",
"code":"49"
},
{
"desc":"Model training uses a large number of labeled images. Therefore, before the model training, add labels to the images that are not labeled. You can add labels to images by",
"product_code":"modelarts",
"title":"Image Classification",
"uri":"modelarts_23_0011.html",
"doc_type":"usermanual",
"p_code":"49",
"code":"50"
},
{
"desc":"Training a model uses a large number of labeled images. Therefore, label images before the model training. You can add labels to images by manual labeling or auto labelin",
"product_code":"modelarts",
"title":"Object Detection",
"uri":"modelarts_23_0012.html",
"doc_type":"usermanual",
"p_code":"49",
"code":"51"
},
{
"desc":"Model training requires a large amount of labeled data. Therefore, before the model training, add labels to the files that are not labeled. In addition, you can modify, d",
"product_code":"modelarts",
"title":"Text Classification",
"uri":"modelarts_23_0013.html",
"doc_type":"usermanual",
"p_code":"49",
"code":"52"
},
{
"desc":"Named entity recognition assigns labels to named entities in text, such as time and locations. Before labeling, you need to understand the following:A label name can cont",
"product_code":"modelarts",
"title":"Named Entity Recognition",
"uri":"modelarts_23_0014.html",
"doc_type":"usermanual",
"p_code":"49",
"code":"53"
},
{
"desc":"Triplet labeling is suitable for scenarios where structured information, such as subjects, predicates, and objects, needs to be labeled in statements. With this function,",
"product_code":"modelarts",
"title":"Text Triplet",
"uri":"modelarts_23_0211.html",
"doc_type":"usermanual",
"p_code":"49",
"code":"54"
},
{
"desc":"Model training requires a large amount of labeled data. Therefore, before the model training, label the unlabeled audio files. ModelArts enables you to label audio files ",
"product_code":"modelarts",
"title":"Sound Classification",
"uri":"modelarts_23_0015.html",
"doc_type":"usermanual",
"p_code":"49",
"code":"55"
},
{
"desc":"Model training requires a large amount of labeled data. Therefore, before the model training, label the unlabeled audio files. ModelArts enables you to label audio files ",
"product_code":"modelarts",
"title":"Speech Labeling",
"uri":"modelarts_23_0016.html",
"doc_type":"usermanual",
"p_code":"49",
"code":"56"
},
{
"desc":"Model training requires a large amount of labeled data. Therefore, before the model training, label the unlabeled audio files. ModelArts enables you to label audio files.",
"product_code":"modelarts",
"title":"Speech Paragraph Labeling",
"uri":"modelarts_23_0017.html",
"doc_type":"usermanual",
"p_code":"49",
"code":"57"
},
{
"desc":"Model training requires a large amount of labeled video data. Therefore, before the model training, label the unlabeled video files. ModelArts enables you to label video ",
"product_code":"modelarts",
"title":"Video Labeling",
"uri":"modelarts_23_0282.html",
"doc_type":"usermanual",
"p_code":"49",
"code":"58"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Importing Data",
"uri":"modelarts_23_0005.html",
"doc_type":"usermanual",
"p_code":"46",
"code":"59"
},
{
"desc":"After a dataset is created, you can directly synchronize data from the dataset. Alternatively, you can import more data by importing the dataset. Data can be imported fro",
"product_code":"modelarts",
"title":"Import Operation",
"uri":"modelarts_23_0006.html",
"doc_type":"usermanual",
"p_code":"59",
"code":"60"
},
{
"desc":"When a dataset is imported, the data storage directory and file name must comply with the ModelArts specifications if the data to be used is stored in OBS.Only the follow",
"product_code":"modelarts",
"title":"Specifications for Importing Data from an OBS Directory",
"uri":"modelarts_23_0008.html",
"doc_type":"usermanual",
"p_code":"59",
"code":"61"
},
{
"desc":"The manifest file defines the mapping between labeling objects and content. The Manifest file import mode means that the manifest file is used for dataset import. The man",
"product_code":"modelarts",
"title":"Specifications for Importing the Manifest File",
"uri":"modelarts_23_0009.html",
"doc_type":"usermanual",
"p_code":"59",
"code":"62"
},
{
"desc":"A dataset includes labeled and unlabeled data. You can select images or filter data based on the filter criteria and export to a new dataset or the specified OBS director",
"product_code":"modelarts",
"title":"Exporting Data",
"uri":"modelarts_23_0214.html",
"doc_type":"usermanual",
"p_code":"46",
"code":"63"
},
{
"desc":"For a created dataset, you can modify its basic information to match service changes.You have created a dataset.Log in to the ModelArts management console. In the left na",
"product_code":"modelarts",
"title":"Modifying a Dataset",
"uri":"modelarts_23_0020.html",
"doc_type":"usermanual",
"p_code":"46",
"code":"64"
},
{
"desc":"ModelArts distinguishes data of the same source according to versions labeled at different time, which facilitates the selection of dataset versions during subsequent mod",
"product_code":"modelarts",
"title":"Publishing a Dataset",
"uri":"modelarts_23_0018.html",
"doc_type":"usermanual",
"p_code":"46",
"code":"65"
},
{
"desc":"If a dataset is no longer in use, you can delete it to release resources.After a dataset is deleted, if you need to delete the data in the dataset input and output paths ",
"product_code":"modelarts",
"title":"Deleting a Dataset",
"uri":"modelarts_23_0021.html",
"doc_type":"usermanual",
"p_code":"46",
"code":"66"
},
{
"desc":"After labeling data, you can publish the dataset to multiple versions for management. For the published versions, you can view the dataset version updates, set the curren",
"product_code":"modelarts",
"title":"Managing Dataset Versions",
"uri":"modelarts_23_0019.html",
"doc_type":"usermanual",
"p_code":"46",
"code":"67"
},
{
"desc":"Only datasets of image classification and object detection types support the auto hard example detection function.In a dataset, labeled or unlabeled image data can be lab",
"product_code":"modelarts",
"title":"Confirming Hard Examples",
"uri":"modelarts_23_0223.html",
"doc_type":"usermanual",
"p_code":"46",
"code":"68"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"DevEnviron (New)",
"uri":"modelarts_30_0000.html",
"doc_type":"usermanual",
"p_code":"",
"code":"69"
},
{
"desc":"This document describes the DevEnviron notebook functions of the new version.Software development is a process of reducing developer costs and improving development exper",
"product_code":"modelarts",
"title":"DevEnviron Overview",
"uri":"modelarts_30_0001.html",
"doc_type":"usermanual",
"p_code":"69",
"code":"70"
},
{
"desc":"ModelArts provides flexible, open development environments. Select a development environment based on site requirements.In-cloud notebook that is out of the box, relievin",
"product_code":"modelarts",
"title":"DevEnviron Application Scenarios",
"uri":"modelarts_30_0002.html",
"doc_type":"usermanual",
"p_code":"69",
"code":"71"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Managing Notebook Instances",
"uri":"modelarts_30_0003.html",
"doc_type":"usermanual",
"p_code":"69",
"code":"72"
},
{
"desc":"Before developing a model, create a notebook instance and access it for coding.Only running notebook instances can be accessed or stopped.A maximum of 10 notebook instanc",
"product_code":"modelarts",
"title":"Creating a Notebook Instance",
"uri":"modelarts_30_0004.html",
"doc_type":"usermanual",
"p_code":"72",
"code":"73"
},
{
"desc":"Access a notebook instance in the Running state for coding.The methods of accessing notebook instances vary depending on the AI engine based on which the instance was cre",
"product_code":"modelarts",
"title":"Accessing a Notebook Instance",
"uri":"modelarts_30_0005.html",
"doc_type":"usermanual",
"p_code":"72",
"code":"74"
},
{
"desc":"Stop the notebook instances that are not needed. You can also restart a stopped instance.Log in to the ModelArts management console. Choose DevEnviron > Notebook in the n",
"product_code":"modelarts",
"title":"Starting, Stopping, or Deleting a Notebook Instance",
"uri":"modelarts_30_0006.html",
"doc_type":"usermanual",
"p_code":"72",
"code":"75"
},
{
"desc":"Storage varies depending on performance, usability, and cost. No storage media can cover all scenarios. Learning about in-cloud storage application scenarios for better u",
"product_code":"modelarts",
"title":"Selecting Storage in DevEnviron",
"uri":"modelarts_30_0033.html",
"doc_type":"usermanual",
"p_code":"72",
"code":"76"
},
{
"desc":"If a notebook instance uses an EVS disk for storage, the disk is mounted to /home/ma-user/work/ of the notebook container and the disk capacity can be expanded by up to 2",
"product_code":"modelarts",
"title":"Dynamically Expanding EVS Disk Capacity",
"uri":"modelarts_30_0040.html",
"doc_type":"usermanual",
"p_code":"72",
"code":"77"
},
{
"desc":"During the creation of a notebook instance, if you set a whitelist for remotely accessing it, you can change the IP addresses in the whitelist on the notebook instance de",
"product_code":"modelarts",
"title":"Changing an IP Address for Remotely Accessing a Notebook Instance",
"uri":"modelarts_30_0023.html",
"doc_type":"usermanual",
"p_code":"72",
"code":"78"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Using JupyterLab to Develop Models",
"uri":"modelarts_30_0007.html",
"doc_type":"usermanual",
"p_code":"69",
"code":"79"
},
{
"desc":"JupyterLab is the next-generation web-based interactive development environment of Jupyter Notebook, enabling you to compile notebooks, operate terminals, edit Markdown t",
"product_code":"modelarts",
"title":"JupyterLab Overview and Common Operations",
"uri":"modelarts_30_0009.html",
"doc_type":"usermanual",
"p_code":"79",
"code":"80"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Uploading Files to JupyterLab",
"uri":"modelarts_30_0041.html",
"doc_type":"usermanual",
"p_code":"79",
"code":"81"
},
{
"desc":"Easy and fast file uploading is a common requirement in AI development.Before the optimization, ModelArts only allowed local files not exceeding 100 MB to be directly upl",
"product_code":"modelarts",
"title":"Scenarios",
"uri":"modelarts_30_0042.html",
"doc_type":"usermanual",
"p_code":"81",
"code":"82"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Uploading Files from a Local Path to JupyterLab",
"uri":"modelarts_30_0043.html",
"doc_type":"usermanual",
"p_code":"81",
"code":"83"
},
{
"desc":"JupyterLab provides multiple methods for uploading files.For a file that does not exceed 100 MB, directly upload it, and details such as the file size, upload progress, a",
"product_code":"modelarts",
"title":"Upload Scenarios and Entries",
"uri":"modelarts_30_0044.html",
"doc_type":"usermanual",
"p_code":"83",
"code":"84"
},
{
"desc":"For a file not exceeding 100 MB, directly upload it to the target notebook instance. Detailed information, such as the file size, upload progress, and upload speed are di",
"product_code":"modelarts",
"title":"Uploading a Local File Less Than 100 MB to JupyterLab",
"uri":"modelarts_30_0045.html",
"doc_type":"usermanual",
"p_code":"83",
"code":"85"
},
{
"desc":"For a file that exceeds 100 MB but does not exceed 5 GB, upload the file to OBS (an object bucket or a parallel file system), and then download the file from OBS to the t",
"product_code":"modelarts",
"title":"Uploading a Local File with a Size Ranging from 100 MB to 5 GB to JupyterLab",
"uri":"modelarts_30_0046.html",
"doc_type":"usermanual",
"p_code":"83",
"code":"86"
},
{
"desc":"A file exceeding 5 GB cannot be directly uploaded to JupyterLab.To upload files exceeding 5 GB, upload them to OBS. Then, call the ModelArts MoXing or SDK API in the targ",
"product_code":"modelarts",
"title":"Uploading a Local File Larger Than 5 GB to JupyterLab",
"uri":"modelarts_30_0047.html",
"doc_type":"usermanual",
"p_code":"83",
"code":"87"
},
{
"desc":"Files can be cloned from a GitHub open-source repository to JupyterLab.Use JupyterLab to open a running notebook instance.Click in the navigation bar on the top of the J",
"product_code":"modelarts",
"title":"Cloning an open-source repository in GitHub",
"uri":"modelarts_30_0048.html",
"doc_type":"usermanual",
"p_code":"81",
"code":"88"
},
{
"desc":"In JupyterLab, you can download files from OBS to a notebook instance.Use JupyterLab to open a running notebook instance.Click in the navigation bar on the top of the Ju",
"product_code":"modelarts",
"title":"Uploading OBS Files to JupyterLab",
"uri":"modelarts_30_0049.html",
"doc_type":"usermanual",
"p_code":"81",
"code":"89"
},
{
"desc":"Files can be downloaded through remote file addresses to JupyterLab.Method: Enter the URL of a remote file in the text box of a browser, and the file is directly download",
"product_code":"modelarts",
"title":"Uploading Remote Files to JupyterLab",
"uri":"modelarts_30_0050.html",
"doc_type":"usermanual",
"p_code":"81",
"code":"90"
},
{
"desc":"Files created in JupyterLab can be downloaded to a local path. The operations for downloading a file are the same, regardless of whether the created notebook instance use",
"product_code":"modelarts",
"title":"Downloading a File from JupyterLab to a Local Path",
"uri":"modelarts_30_0011.html",
"doc_type":"usermanual",
"p_code":"79",
"code":"91"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"JupyterLab Plug-ins",
"uri":"devtool-modelarts_0211.html",
"doc_type":"usermanual",
"p_code":"79",
"code":"92"
},
{
"desc":"The code parametrization plug-in simplifies notebook cases. You can quickly adjust parameters and train models based on notebook cases without complex code. This plug-in ",
"product_code":"modelarts",
"title":"Code Parametrization Plug-in",
"uri":"devtool-modelarts_0212.html",
"doc_type":"usermanual",
"p_code":"92",
"code":"93"
},
{
"desc":"Notebook instances allow you to use ModelArts SDK to manage OBS, training jobs, models, and real-time services.Your notebook instances have automatically obtained your AK",
"product_code":"modelarts",
"title":"Using ModelArts SDK",
"uri":"modelarts_30_0030.html",
"doc_type":"usermanual",
"p_code":"79",
"code":"94"
},
{
"desc":"This section describes how to use PuTTY to remotely log in to a notebook instance on the cloud in the Windows environment.You have created a notebook instance with remote",
"product_code":"modelarts",
"title":"Configuring a Local IDE Accessed Using SSH",
"uri":"modelarts_30_0038.html",
"doc_type":"usermanual",
"p_code":"69",
"code":"95"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"DevEnviron (Old)",
"uri":"modelarts_23_0032.html",
"doc_type":"usermanual",
"p_code":"",
"code":"96"
},
{
"desc":"ModelArts integrates the open-source Jupyter Notebook and JupyterLab to provide you with online interactive development and debugging environments. You can use the Notebo",
"product_code":"modelarts",
"title":"Introduction to Notebook",
"uri":"modelarts_23_0033.html",
"doc_type":"usermanual",
"p_code":"96",
"code":"97"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Managing Notebook Instances",
"uri":"modelarts_23_0111.html",
"doc_type":"usermanual",
"p_code":"96",
"code":"98"
},
{
"desc":"Before developing a model, create a notebook instance, open it, and perform encoding.Only notebook instances in the Running state can be started.A maximum of 10 notebook ",
"product_code":"modelarts",
"title":"Creating a Notebook Instance",
"uri":"modelarts_23_0034.html",
"doc_type":"usermanual",
"p_code":"98",
"code":"99"
},
{
"desc":"You can open a created notebook instance (that is, an instance in the Running state) and start coding in the development environment.Go to the Jupyter Notebook page.In th",
"product_code":"modelarts",
"title":"Opening a Notebook Instance",
"uri":"modelarts_23_0325.html",
"doc_type":"usermanual",
"p_code":"98",
"code":"100"
},
{
"desc":"You can stop unwanted notebook instances to prevent unnecessary fees. You can also start a notebook instance that is in the Stopped state to use it again.Log in to the Mo",
"product_code":"modelarts",
"title":"Starting or Stopping a Notebook Instance",
"uri":"modelarts_23_0041.html",
"doc_type":"usermanual",
"p_code":"98",
"code":"101"
},
{
"desc":"You can delete notebook instances that are no longer used to release resources.Log in to the ModelArts management console. In the left navigation pane, choose DevEnviron ",
"product_code":"modelarts",
"title":"Deleting a Notebook Instance",
"uri":"modelarts_23_0042.html",
"doc_type":"usermanual",
"p_code":"98",
"code":"102"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Using Jupyter Notebook",
"uri":"modelarts_23_0035.html",
"doc_type":"usermanual",
"p_code":"96",
"code":"103"
},
{
"desc":"Jupyter Notebook is a web-based application for interactive computing. It can be applied to full-process computing: development, documentation, running code, and presenti",
"product_code":"modelarts",
"title":"Introduction to Jupyter Notebook",
"uri":"modelarts_23_0326.html",
"doc_type":"usermanual",
"p_code":"103",
"code":"104"
},
{
"desc":"This section describes common operations on Jupyter Notebook.In the notebook instance list, locate the row where the target notebook instance resides and click Open in th",
"product_code":"modelarts",
"title":"Common Operations on Jupyter Notebook",
"uri":"modelarts_23_0120.html",
"doc_type":"usermanual",
"p_code":"103",
"code":"105"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Configuring the Jupyter Notebook Environment",
"uri":"modelarts_23_0327.html",
"doc_type":"usermanual",
"p_code":"103",
"code":"106"
},
{
"desc":"For developers who are used to coding, the terminal function is very convenient and practical. This section describes how to enable the terminal function in a notebook in",
"product_code":"modelarts",
"title":"Using the Notebook Terminal Function",
"uri":"modelarts_23_0117.html",
"doc_type":"usermanual",
"p_code":"106",
"code":"107"
},
{
"desc":"For a GPU-based notebook instance, you can switch different versions of CUDA on the Terminal page of Jupyter.CPU-based notebook instances do not use CUDA. Therefore, the ",
"product_code":"modelarts",
"title":"Switching the CUDA Version on the Terminal Page of a GPU-based Notebook Instance",
"uri":"modelarts_23_0280.html",
"doc_type":"usermanual",
"p_code":"106",
"code":"108"
},
{
"desc":"Multiple environments have been installed in ModelArts notebook instances, including TensorFlow. You can use pip install to install external libraries from a Jupyter note",
"product_code":"modelarts",
"title":"Installing External Libraries and Kernels in Notebook Instances",
"uri":"modelarts_23_0040.html",
"doc_type":"usermanual",
"p_code":"106",
"code":"109"
},
{
"desc":"In notebook instances, you can use ModelArts SDKs to manage OBS, training jobs, models, and real-time services.For details about how to use ModelArts SDKs, see ModelArts ",
"product_code":"modelarts",
"title":"Using ModelArts SDKs",
"uri":"modelarts_23_0039.html",
"doc_type":"usermanual",
"p_code":"103",
"code":"110"
},
{
"desc":"If you specify Storage Path during notebook instance creation, your compiled code will be automatically stored in your specified OBS bucket. If code invocation among diff",
"product_code":"modelarts",
"title":"Synchronizing Files with OBS",
"uri":"modelarts_23_0038.html",
"doc_type":"usermanual",
"p_code":"103",
"code":"111"
},
{
"desc":"After code compiling is finished, you can save the entered code as a .py file which can be used for starting training jobs.Create and open a notebook instance or open an ",
"product_code":"modelarts",
"title":"Using the Convert to Python File Function",
"uri":"modelarts_23_0037.html",
"doc_type":"usermanual",
"p_code":"103",
"code":"112"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Using JupyterLab",
"uri":"modelarts_23_0330.html",
"doc_type":"usermanual",
"p_code":"96",
"code":"113"
},
{
"desc":"JupyterLab is an interactive development environment. It is a next-generation product of Jupyter Notebook. JupyterLab enables you to compile notebooks, operate terminals,",
"product_code":"modelarts",
"title":"Introduction to JupyterLab and Common Operations",
"uri":"modelarts_23_0209.html",
"doc_type":"usermanual",
"p_code":"113",
"code":"114"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Uploading and Downloading Data",
"uri":"modelarts_23_0331.html",
"doc_type":"usermanual",
"p_code":"113",
"code":"115"
},
{
"desc":"On the JupyterLab page, click Upload Files to upload a file. For details, see Uploading a File in Introduction to JupyterLab and Common Operations. If a message is displa",
"product_code":"modelarts",
"title":"Uploading Data to JupyterLab",
"uri":"modelarts_23_0332.html",
"doc_type":"usermanual",
"p_code":"115",
"code":"116"
},
{
"desc":"Only files within 100 MB in JupyterLab can be downloaded to a local PC. You can perform operations in different scenarios based on the storage location selected when crea",
"product_code":"modelarts",
"title":"Downloading a File from JupyterLab",
"uri":"modelarts_23_0333.html",
"doc_type":"usermanual",
"p_code":"115",
"code":"117"
},
{
"desc":"In notebook instances, you can use ModelArts SDKs to manage OBS, training jobs, models, and real-time services.For details about how to use ModelArts SDKs, see ModelArts ",
"product_code":"modelarts",
"title":"Using ModelArts SDKs",
"uri":"modelarts_23_0335.html",
"doc_type":"usermanual",
"p_code":"113",
"code":"118"
},
{
"desc":"If you specify Storage Path during notebook instance creation, your compiled code will be automatically stored in your specified OBS bucket. If code invocation among diff",
"product_code":"modelarts",
"title":"Synchronizing Files with OBS",
"uri":"modelarts_23_0336.html",
"doc_type":"usermanual",
"p_code":"113",
"code":"119"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Training Management (New Version)",
"uri":"modelarts_23_0284.html",
"doc_type":"usermanual",
"p_code":"",
"code":"120"
},
{
"desc":"ModelArts provides model training of both the new and old versions. The new version features enhanced functions, optimized scheduling, and improved APIs. You are advised ",
"product_code":"modelarts",
"title":"Introduction to Model Training",
"uri":"modelarts_23_0285.html",
"doc_type":"usermanual",
"p_code":"120",
"code":"121"
},
{
"desc":"ModelArts uses OBS to store data, and backs up and takes snapshots for models, achieving secure, reliable storage at low costs.OBSObtaining Training DataOBS provides stab",
"product_code":"modelarts",
"title":"Preparing Data",
"uri":"modelarts_23_0351.html",
"doc_type":"usermanual",
"p_code":"120",
"code":"122"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Selecting an Algorithm",
"uri":"modelarts_23_0230.html",
"doc_type":"usermanual",
"p_code":"120",
"code":"123"
},
{
"desc":"Machine learning explores general rules from limited volume of data and uses these rules to predict unknown data. To obtain more accurate prediction results, select a pro",
"product_code":"modelarts",
"title":"Introduction to Algorithm Selection",
"uri":"modelarts_23_0234.html",
"doc_type":"usermanual",
"p_code":"123",
"code":"124"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Using a Custom Script",
"uri":"modelarts_23_0231.html",
"doc_type":"usermanual",
"p_code":"123",
"code":"125"
},
{
"desc":"If the subscribed algorithms cannot meet your requirements or you want to migrate local algorithms to ModelArts for training, use the ModelArts built-in training engines ",
"product_code":"modelarts",
"title":"Introduction to Custom Script",
"uri":"modelarts_23_0283.html",
"doc_type":"usermanual",
"p_code":"125",
"code":"126"
},
{
"desc":"Before you use a custom script to create an algorithm, develop the algorithm code. This section describes how to modify local code for model training on ModelArts.When cr",
"product_code":"modelarts",
"title":"Developing a Custom Script",
"uri":"modelarts_23_0240_0.html",
"doc_type":"usermanual",
"p_code":"125",
"code":"127"
},
{
"desc":"Your locally developed algorithms or algorithms developed using other tools can be uploaded to ModelArts for unified management. Note the following when creating a custom",
"product_code":"modelarts",
"title":"Creating an Algorithm",
"uri":"modelarts_23_0233.html",
"doc_type":"usermanual",
"p_code":"125",
"code":"128"
},
{
"desc":"The preset images can be used in most training scenarios. In certain scenarios, ModelArts allows you to create custom images to train models. Custom images can be used to",
"product_code":"modelarts",
"title":"Using Custom Images",
"uri":"develop-modelarts-0077.html",
"doc_type":"usermanual",
"p_code":"123",
"code":"129"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Performing a Training",
"uri":"modelarts_23_0352.html",
"doc_type":"usermanual",
"p_code":"120",
"code":"130"
},
{
"desc":"ModelArts training management enables you to create training jobs, view training statuses, and manage job versions. Model training is an iterative optimization process. T",
"product_code":"modelarts",
"title":"Creating a Training Job",
"uri":"modelarts_23_0286.html",
"doc_type":"usermanual",
"p_code":"130",
"code":"131"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Viewing Job Details",
"uri":"modelarts_23_0288.html",
"doc_type":"usermanual",
"p_code":"130",
"code":"132"
},
{
"desc":"Log in to the ModelArts management console. In the left navigation pane, choose Training Management > Training Jobs (New). The training job list is displayed by default.C",
"product_code":"modelarts",
"title":"Training Job Details",
"uri":"modelarts_23_0400.html",
"doc_type":"usermanual",
"p_code":"132",
"code":"133"
},
{
"desc":"Any key event of a training job will be recorded at the backend after the training job is displayed for you. You can check events on the training job details page.This he",
"product_code":"modelarts",
"title":"Training Job Event",
"uri":"develop-modelarts-0081.html",
"doc_type":"usermanual",
"p_code":"132",
"code":"134"
},
{
"desc":"On the training job details page, you can preview logs, download logs, search for logs by keyword, and identify training faults in the log pane.Preview logs.You can previ",
"product_code":"modelarts",
"title":"Training Log Details",
"uri":"modelarts_23_0401.html",
"doc_type":"usermanual",
"p_code":"132",
"code":"135"
},
{
"desc":"In the Resource Usages pane, view resource usage of compute nodes.Operation 1: If a training job uses multiple compute nodes, choose a node from the drop-down list box to",
"product_code":"modelarts",
"title":"Resource Usage",
"uri":"modelarts_23_0402.html",
"doc_type":"usermanual",
"p_code":"132",
"code":"136"
},
{
"desc":"In the training job list, click Stop in the Operation column of a training job that is in creating, pending, or running state to stop the job.A training job in completed,",
"product_code":"modelarts",
"title":"Stopping, Rebuilding, or Searching for a Training Job",
"uri":"modelarts_23_0287.html",
"doc_type":"usermanual",
"p_code":"130",
"code":"137"
},
{
"desc":"Release resources of a training job when not in use.On the Training Jobs page, click Delete in the Operation column of the target training job.Go to OBS and delete the OB",
"product_code":"modelarts",
"title":"Releasing Training Job Resources",
"uri":"modelarts_23_0353.html",
"doc_type":"usermanual",
"p_code":"130",
"code":"138"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Training Hyperparameter Search",
"uri":"modelarts_23_0289.html",
"doc_type":"usermanual",
"p_code":"120",
"code":"139"
},
{
"desc":"The new version of ModelArts training jobs supports hyperparameter search, which can automatically search for optimal hyperparameters for your models.During model trainin",
"product_code":"modelarts",
"title":"Introduction to Hyperparameter Search",
"uri":"modelarts_23_0290.html",
"doc_type":"usermanual",
"p_code":"139",
"code":"140"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Search Algorithm",
"uri":"modelarts_23_0296.html",
"doc_type":"usermanual",
"p_code":"139",
"code":"141"
},
{
"desc":"In Bayesian optimization, it is assumed that there exists a functional relationship between hyperparameters and the objective function. Based on the evaluation values of ",
"product_code":"modelarts",
"title":"Bayesian Optimization (SMAC)",
"uri":"modelarts_23_0297.html",
"doc_type":"usermanual",
"p_code":"141",
"code":"142"
},
{
"desc":"The tree-structured parzen estimator (TPE) algorithm uses the Gaussian mixture model to learn the model hyperparameters. On each trial, for each parameter, TPE fits one G",
"product_code":"modelarts",
"title":"TPE Algorithm",
"uri":"modelarts_23_0303_0.html",
"doc_type":"usermanual",
"p_code":"141",
"code":"143"
},
{
"desc":"The simulated annealing algorithm is a simple but effective variant on random search that leverages smoothness in the response surface. The annealing rate is not adaptive",
"product_code":"modelarts",
"title":"Simulated Annealing Algorithm",
"uri":"modelarts_23_0304_0.html",
"doc_type":"usermanual",
"p_code":"141",
"code":"144"
},
{
"desc":"Hyperparameters that you want to optimize need to be defined when you configure Hyperparameters. You can specify the name, type, default value, and constraints. For detai",
"product_code":"modelarts",
"title":"Creating a Hyperparameter Search Job",
"uri":"modelarts_23_0302_0.html",
"doc_type":"usermanual",
"p_code":"139",
"code":"145"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"AI Application Management",
"uri":"modelarts_23_0051.html",
"doc_type":"usermanual",
"p_code":"",
"code":"146"
},
{
"desc":"AI development and optimization require frequent iterations and debugging. Changes in datasets, training code, or parameters affect the quality of models. If the metadata",
"product_code":"modelarts",
"title":"Introduction to AI Application Management",
"uri":"modelarts_23_0052.html",
"doc_type":"usermanual",
"p_code":"146",
"code":"147"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Creating an AI Application",
"uri":"modelarts_23_0204.html",
"doc_type":"usermanual",
"p_code":"146",
"code":"148"
},
{
"desc":"You can create a training job on ModelArts and perform training to obtain a satisfactory model. Then import the model to Model Management for centralized management. In a",
"product_code":"modelarts",
"title":"Importing a Meta Model from a Training Job",
"uri":"modelarts_23_0054.html",
"doc_type":"usermanual",
"p_code":"148",
"code":"149"
},
{
"desc":"Because the configurations of models with the same functions are similar, ModelArts integrates the configurations of such models into a common template. By using this tem",
"product_code":"modelarts",
"title":"Importing a Meta Model from a Template",
"uri":"modelarts_23_0205.html",
"doc_type":"usermanual",
"p_code":"148",
"code":"150"
},
{
"desc":"For AI engines that are not supported by ModelArts, you can import the models you compile to ModelArts from custom images.For details about the specifications and descrip",
"product_code":"modelarts",
"title":"Importing a Meta Model from a Container Image",
"uri":"modelarts_23_0206.html",
"doc_type":"usermanual",
"p_code":"148",
"code":"151"
},
{
"desc":"In scenarios where frequently-used frameworks are used for model development and training, you can import the model to ModelArts and use it to create an AI application fo",
"product_code":"modelarts",
"title":"Importing a Meta Model from OBS",
"uri":"modelarts_23_0207.html",
"doc_type":"usermanual",
"p_code":"148",
"code":"152"
},
{
"desc":"To facilitate source tracing and repeated AI application tuning, ModelArts provides the AI application version management function. You can manage models based on version",
"product_code":"modelarts",
"title":"Managing AI Application Versions",
"uri":"modelarts_23_0055.html",
"doc_type":"usermanual",
"p_code":"146",
"code":"153"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Deploying a Service",
"uri":"modelarts_23_0057.html",
"doc_type":"usermanual",
"p_code":"",
"code":"154"
},
{
"desc":"After a training job is complete and an AI application is generated, you can deploy the model on the Service Deployment page. You can also deploy the model imported from ",
"product_code":"modelarts",
"title":"Deploying the AI Applications as Services",
"uri":"modelarts_23_0058.html",
"doc_type":"usermanual",
"p_code":"154",
"code":"155"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Real-Time Services",
"uri":"modelarts_23_0059.html",
"doc_type":"usermanual",
"p_code":"154",
"code":"156"
},
{
"desc":"After an AI application is prepared, you can deploy the AI application as a real-time service and predict and call the service.A maximum of 20 real-time services can be d",
"product_code":"modelarts",
"title":"Deploying a Model as a Real-Time Service",
"uri":"modelarts_23_0060.html",
"doc_type":"usermanual",
"p_code":"156",
"code":"157"
},
{
"desc":"After an AI application is deployed as a real-time service, you can access the service page to view its details.Log in to the ModelArts management console and choose Serv",
"product_code":"modelarts",
"title":"Viewing Service Details",
"uri":"modelarts_23_0061.html",
"doc_type":"usermanual",
"p_code":"156",
"code":"158"
},
{
"desc":"After an AI application is deployed as a real-time service, you can debug code or add files for testing on the Prediction tab page. Based on the input request (JSON text ",
"product_code":"modelarts",
"title":"Testing a Service",
"uri":"modelarts_23_0062.html",
"doc_type":"usermanual",
"p_code":"156",
"code":"159"
},
{
"desc":"If a real-time service is in the Running state, the real-time service has been deployed successfully. This service provides a standard RESTful API for users to call. Befo",
"product_code":"modelarts",
"title":"Accessing a Real-Time Service (Token-based Authentication)",
"uri":"modelarts_23_0063.html",
"doc_type":"usermanual",
"p_code":"156",
"code":"160"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Batch Services",
"uri":"modelarts_23_0065.html",
"doc_type":"usermanual",
"p_code":"154",
"code":"161"
},
{
"desc":"After an AI application is prepared, you can deploy it as a batch service. The Service Deployment > Batch Services page lists all batch services. You can enter a service ",
"product_code":"modelarts",
"title":"Deploying a Model as a Batch Service",
"uri":"modelarts_23_0066.html",
"doc_type":"usermanual",
"p_code":"161",
"code":"162"
},
{
"desc":"When deploying a batch service, you can select the location of the output data directory. You can view the running result of the batch service that is in the Running comp",
"product_code":"modelarts",
"title":"Viewing the Batch Service Prediction Result",
"uri":"modelarts_23_0067.html",
"doc_type":"usermanual",
"p_code":"161",
"code":"163"
},
{
"desc":"For a deployed service, you can modify its basic information to match service changes and upgrade it. You can modify the basic information about a service in either of th",
"product_code":"modelarts",
"title":"Upgrading a Service",
"uri":"modelarts_23_0071.html",
"doc_type":"usermanual",
"p_code":"154",
"code":"164"
},
{
"desc":"You can start services in the Successful, Abnormal, or Stopped status. Services in the Deploying state cannot be started. You can start a service in the following ways:Lo",
"product_code":"modelarts",
"title":"Starting or Stopping a Service",
"uri":"modelarts_23_0072.html",
"doc_type":"usermanual",
"p_code":"154",
"code":"165"
},
{
"desc":"If a service is no longer in use, you can delete it to release resources.Log in to the ModelArts management console and choose Service Deployment from the left navigation",
"product_code":"modelarts",
"title":"Deleting a Service",
"uri":"modelarts_23_0073.html",
"doc_type":"usermanual",
"p_code":"154",
"code":"166"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Model Package Specifications",
"uri":"modelarts_23_0090.html",
"doc_type":"usermanual",
"p_code":"",
"code":"167"
},
{
"desc":"When creating an AI application on the AI application management page, make sure that any meta model imported from OBS complies with certain specifications.The model pack",
"product_code":"modelarts",
"title":"Model Package Specifications",
"uri":"modelarts_23_0091.html",
"doc_type":"usermanual",
"p_code":"167",
"code":"168"
},
{
"desc":"A model developer needs to compile a configuration file when publishing a model. The model configuration file describes the model usage, computing framework, precision, i",
"product_code":"modelarts",
"title":"Specifications for Compiling the Model Configuration File",
"uri":"modelarts_23_0092.html",
"doc_type":"usermanual",
"p_code":"167",
"code":"169"
},
{
"desc":"This section describes the general method of editing model inference code in ModelArts. This section also provides an inference code example for the TensorFlow engine and",
"product_code":"modelarts",
"title":"Specifications for Writing Model Inference Code",
"uri":"modelarts_23_0093.html",
"doc_type":"usermanual",
"p_code":"167",
"code":"170"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Model Templates",
"uri":"modelarts_23_0097.html",
"doc_type":"usermanual",
"p_code":"",
"code":"171"
},
{
"desc":"Because the configurations of models with the same functions are similar, ModelArts integrates the configurations of such models into a common template. By using this tem",
"product_code":"modelarts",
"title":"Introduction to Model Templates",
"uri":"modelarts_23_0098.html",
"doc_type":"usermanual",
"p_code":"171",
"code":"172"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Templates",
"uri":"modelarts_23_0118.html",
"doc_type":"usermanual",
"p_code":"171",
"code":"173"
},
{
"desc":"AI engine: TensorFlow 1.8; Environment: Python 2.7; Input and output mode: undefined mode. Select an appropriate input and output mode based on the model function or appl",
"product_code":"modelarts",
"title":"TensorFlow-py27 General Template",
"uri":"modelarts_23_0161.html",
"doc_type":"usermanual",
"p_code":"173",
"code":"174"
},
{
"desc":"AI engine: TensorFlow 1.8; Environment: Python 3.6; Input and output mode: Undefined. Select an appropriate input and output mode based on the model function or applicati",
"product_code":"modelarts",
"title":"TensorFlow-py36 General Template",
"uri":"modelarts_23_0262.html",
"doc_type":"usermanual",
"p_code":"173",
"code":"175"
},
{
"desc":"AI engine: MXNet 1.2.1; Environment: Python 2.7; Input and output mode: undefined mode. Select an appropriate input and output mode based on the model function or applica",
"product_code":"modelarts",
"title":"MXNet-py27 General Template",
"uri":"modelarts_23_0163.html",
"doc_type":"usermanual",
"p_code":"173",
"code":"176"
},
{
"desc":"AI engine: MXNet 1.2.1; Environment: Python 3.7; Input and output mode: undefined mode. Select an appropriate input and output mode based on the model function or applica",
"product_code":"modelarts",
"title":"MXNet-py37 General Template",
"uri":"modelarts_23_0164.html",
"doc_type":"usermanual",
"p_code":"173",
"code":"177"
},
{
"desc":"AI engine: PyTorch 1.0; Environment: Python 2.7; Input and output mode: Undefined. Select an appropriate input and output mode based on the model function or application ",
"product_code":"modelarts",
"title":"PyTorch-py27 General Template",
"uri":"modelarts_23_0165.html",
"doc_type":"usermanual",
"p_code":"173",
"code":"178"
},
{
"desc":"AI engine: PyTorch 1.0; Environment: Python 3.7; Input and output mode: undefined mode. Select an appropriate input and output mode based on the model function or applica",
"product_code":"modelarts",
"title":"PyTorch-py37 General Template",
"uri":"modelarts_23_0166.html",
"doc_type":"usermanual",
"p_code":"173",
"code":"179"
},
{
"desc":"AI engine: CPU-based Caffe 1.0; Environment: Python 2.7; Input and output mode: undefined mode. Select an appropriate input and output mode based on the model function or",
"product_code":"modelarts",
"title":"Caffe-CPU-py27 General Template",
"uri":"modelarts_23_0167.html",
"doc_type":"usermanual",
"p_code":"173",
"code":"180"
},
{
"desc":"AI engine: GPU-based Caffe 1.0; Environment: Python 2.7; Input and output mode: Undefined. Select an appropriate input and output mode based on the model function or appl",
"product_code":"modelarts",
"title":"Caffe-GPU-py27 General Template",
"uri":"modelarts_23_0168.html",
"doc_type":"usermanual",
"p_code":"173",
"code":"181"
},
{
"desc":"AI engine: CPU-based Caffe 1.0; Environment: Python 3.7; Input and output mode: undefined mode. Select an appropriate input and output mode based on the model function or",
"product_code":"modelarts",
"title":"Caffe-CPU-py37 General Template",
"uri":"modelarts_23_0169.html",
"doc_type":"usermanual",
"p_code":"173",
"code":"182"
},
{
"desc":"AI engine: GPU-based Caffe 1.0; Environment: Python 3.7; Input and output mode: undefined mode. Select an appropriate input and output mode based on the model function or",
"product_code":"modelarts",
"title":"Caffe-GPU-py37 General Template",
"uri":"modelarts_23_0170.html",
"doc_type":"usermanual",
"p_code":"173",
"code":"183"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Input and Output Modes",
"uri":"modelarts_23_0099.html",
"doc_type":"usermanual",
"p_code":"171",
"code":"184"
},
{
"desc":"This is a built-in input and output mode for object detection. The models using this mode are identified as object detection models. The prediction request path is /, the",
"product_code":"modelarts",
"title":"Built-in Object Detection Mode",
"uri":"modelarts_23_0100.html",
"doc_type":"usermanual",
"p_code":"184",
"code":"185"
},
{
"desc":"The built-in image processing input and output mode can be applied to models such as image classification, object detection, and image semantic segmentation. The predicti",
"product_code":"modelarts",
"title":"Built-in Image Processing Mode",
"uri":"modelarts_23_0101.html",
"doc_type":"usermanual",
"p_code":"184",
"code":"186"
},
{
"desc":"This is a built-in input and output mode for predictive analytics. The models using this mode are identified as predictive analytics models. The prediction request path i",
"product_code":"modelarts",
"title":"Built-in Predictive Analytics Mode",
"uri":"modelarts_23_0102.html",
"doc_type":"usermanual",
"p_code":"184",
"code":"187"
},
{
"desc":"The undefined mode does not define the input and output mode. The input and output mode is determined by the model. Select this mode only when the existing input and outp",
"product_code":"modelarts",
"title":"Undefined Mode",
"uri":"modelarts_23_0103.html",
"doc_type":"usermanual",
"p_code":"184",
"code":"188"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Examples of Custom Scripts",
"uri":"modelarts_23_0172.html",
"doc_type":"usermanual",
"p_code":"",
"code":"189"
},
{
"desc":"TensorFlow has two types of APIs: Keras and tf. Keras and tf use different code for training and saving models, but the same code for inference.Inference code must be inh",
"product_code":"modelarts",
"title":"TensorFlow",
"uri":"modelarts_23_0173.html",
"doc_type":"usermanual",
"p_code":"189",
"code":"190"
},
{
"desc":"Inference code must be inherited from the BaseService class. For details about the import statements of different types of parent model classes, see Table 1.",
"product_code":"modelarts",
"title":"TensorFlow 2.1",
"uri":"modelarts_23_0301.html",
"doc_type":"usermanual",
"p_code":"189",
"code":"191"
},
{
"desc":"Inference code must be inherited from the BaseService class. For details about the import statements of different types of parent model classes, see Table 1.",
"product_code":"modelarts",
"title":"PyTorch",
"uri":"modelarts_23_0175.html",
"doc_type":"usermanual",
"p_code":"189",
"code":"192"
},
{
"desc":"lenet_train_test.prototxt filelenet_solver.prototxt fileTrain the model.The caffemodel file is generated after model training. Rewrite the lenet_train_test.prototxt file ",
"product_code":"modelarts",
"title":"Caffe",
"uri":"modelarts_23_0176.html",
"doc_type":"usermanual",
"p_code":"189",
"code":"193"
},
{
"desc":"Before training, download the iris.csv dataset, decompress it, and upload it to the /home/ma-user/work/ directory of the notebook instance. Download the iris.csv dataset ",
"product_code":"modelarts",
"title":"XGBoost",
"uri":"modelarts_23_0177.html",
"doc_type":"usermanual",
"p_code":"189",
"code":"194"
},
{
"desc":"After the model is saved, it must be uploaded to the OBS directory before being published. The config.json configuration and the customize_service.py inference code must ",
"product_code":"modelarts",
"title":"Spark",
"uri":"modelarts_23_0178.html",
"doc_type":"usermanual",
"p_code":"189",
"code":"195"
},
{
"desc":"Before training, download the iris.csv dataset, decompress it, and upload it to the /home/ma-user/work/ directory of the notebook instance. Download the iris.csv dataset ",
"product_code":"modelarts",
"title":"Scikit Learn",
"uri":"modelarts_23_0179.html",
"doc_type":"usermanual",
"p_code":"189",
"code":"196"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Using Custom Images",
"uri":"en-us_topic_0000001799497932.html",
"doc_type":"usermanual",
"p_code":"",
"code":"197"
},
{
"desc":"Frequently-used images are preset in ModelArts. However, if you have special requirements for the deep learning engine or development library, the preset images cannot me",
"product_code":"modelarts",
"title":"Introduction to Custom Images",
"uri":"modelarts_23_0084.html",
"doc_type":"usermanual",
"p_code":"197",
"code":"198"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Using a Custom Image to Train Models (New-Version Training)",
"uri":"docker-modelarts_0017.html",
"doc_type":"usermanual",
"p_code":"197",
"code":"199"
},
{
"desc":"When you use a locally developed model or training script to create a custom image, ensure that the custom image complies with the specifications defined by ModelArts.A c",
"product_code":"modelarts",
"title":"Specifications for Custom Images Used for Training Jobs",
"uri":"develop-modelarts-0079.html",
"doc_type":"usermanual",
"p_code":"199",
"code":"200"
},
{
"desc":"To migrate an image to the new training management version, perform the following operations:Add the default user group ma-group (GID = 100) for the image of the new-vers",
"product_code":"modelarts",
"title":"Migrating an Image to ModelArts Training",
"uri":"docker-modelarts_0029.html",
"doc_type":"usermanual",
"p_code":"199",
"code":"201"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Using a Custom Image to Create AI applications",
"uri":"modelarts_23_0218.html",
"doc_type":"usermanual",
"p_code":"197",
"code":"202"
},
{
"desc":"When creating an image using locally developed models, ensure that they meet the specifications defined by ModelArts.Custom images cannot contain malicious code.A custom ",
"product_code":"modelarts",
"title":"Custom Image Specifications for Creating an AI Application",
"uri":"modelarts_23_0219.html",
"doc_type":"usermanual",
"p_code":"202",
"code":"203"
},
{
"desc":"For AI engines that are not supported by ModelArts, you can import the models you compile to ModelArts from custom images for creating the AI applications. This section d",
"product_code":"modelarts",
"title":"Deploying an AI Application Created Using a Custom Image as a Service",
"uri":"modelarts_23_0270.html",
"doc_type":"usermanual",
"p_code":"202",
"code":"204"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"FAQ",
"uri":"docker-modelarts_0016.html",
"doc_type":"usermanual",
"p_code":"197",
"code":"205"
},
{
"desc":"This section describes how to upload images to SWR.Log in to the SWR console.Click Create Organization in the upper right corner and enter an organization name to create ",
"product_code":"modelarts",
"title":"How Can I Upload Images to SWR?",
"uri":"docker-modelarts_0018.html",
"doc_type":"usermanual",
"p_code":"205",
"code":"206"
},
{
"desc":"In a Dockerfile, use the ENV instruction to configure environment variables. For details, see Dockerfile reference.",
"product_code":"modelarts",
"title":"How Do I Configure Environment Variables for an Image?",
"uri":"docker-modelarts_0019.html",
"doc_type":"usermanual",
"p_code":"205",
"code":"207"
},
{
"desc":"When using ModelArts for full-process AI development, you can use two different resource pools.Public Resource Pool: provides public large-scale computing clusters, which",
"product_code":"modelarts",
"title":"Resource Pools",
"uri":"modelarts_23_0076.html",
"doc_type":"usermanual",
"p_code":"",
"code":"208"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Permissions Management",
"uri":"modelarts_23_0077.html",
"doc_type":"usermanual",
"p_code":"",
"code":"209"
},
{
"desc":"A fine-grained policy is a set of permissions defining which operations on which cloud services can be performed. Each policy can define multiple permissions. After a pol",
"product_code":"modelarts",
"title":"Fine-grained Policy",
"uri":"modelarts_23_0078.html",
"doc_type":"usermanual",
"p_code":"209",
"code":"210"
},
{
"desc":"A fine-grained policy consists of the policy version (the Version field) and statement (the Statement field).Version: Distinguishes between role-based access control (RBA",
"product_code":"modelarts",
"title":"Policy Language",
"uri":"modelarts_23_0079.html",
"doc_type":"usermanual",
"p_code":"209",
"code":"211"
},
{
"desc":"If default policies cannot meet the requirements on fine-grained access control, you can create custom policies and assign the policies to the user group.You can create c",
"product_code":"modelarts",
"title":"Creating a Custom Policy",
"uri":"modelarts_23_0080.html",
"doc_type":"usermanual",
"p_code":"209",
"code":"212"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Monitoring",
"uri":"modelarts_23_0186.html",
"doc_type":"usermanual",
"p_code":"",
"code":"213"
},
{
"desc":"The cloud service platform provides Cloud Eye to help you better understand the status of your ModelArts real-time services and models. You can use Cloud Eye to automatic",
"product_code":"modelarts",
"title":"ModelArts Metrics",
"uri":"modelarts_23_0187.html",
"doc_type":"usermanual",
"p_code":"213",
"code":"214"
},
{
"desc":"Setting alarm rules allows you to customize the monitored objects and notification policies so that you can know the status of ModelArts real-time services and models in ",
"product_code":"modelarts",
"title":"Setting Alarm Rules",
"uri":"modelarts_23_0188.html",
"doc_type":"usermanual",
"p_code":"213",
"code":"215"
},
{
"desc":"Cloud Eye on the cloud service platform monitors the status of ModelArts real-time services and model loads. You can obtain the monitoring metrics of each ModelArts real-",
"product_code":"modelarts",
"title":"Viewing Monitoring Metrics",
"uri":"modelarts_23_0189.html",
"doc_type":"usermanual",
"p_code":"213",
"code":"216"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Audit Logs",
"uri":"modelarts_23_0249.html",
"doc_type":"usermanual",
"p_code":"",
"code":"217"
},
{
"desc":"With CTS, you can record operations associated with ModelArts for later query, audit, and backtrack operations.CTS has been enabled. For details, see Enabling CTS",
"product_code":"modelarts",
"title":"Key Operations Recorded by CTS",
"uri":"modelarts_23_0250.html",
"doc_type":"usermanual",
"p_code":"217",
"code":"218"
},
{
"desc":"After CTS is enabled, CTS starts recording operations related to ModelArts. The CTS management console stores the last seven days of operation records. This section descr",
"product_code":"modelarts",
"title":"Viewing Audit Logs",
"uri":"modelarts_23_0251.html",
"doc_type":"usermanual",
"p_code":"217",
"code":"219"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"FAQs",
"uri":"modelarts_05_0000.html",
"doc_type":"usermanual",
"p_code":"",
"code":"220"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"General Issues",
"uri":"modelarts_05_0014.html",
"doc_type":"usermanual",
"p_code":"220",
"code":"221"
},
{
"desc":"ModelArts is a one-stop AI development platform geared toward developers and data scientists of all skill levels. It enables you to rapidly build, train, and deploy model",
"product_code":"modelarts",
"title":"What Is ModelArts?",
"uri":"modelarts_05_0001.html",
"doc_type":"usermanual",
"p_code":"221",
"code":"222"
},
{
"desc":"ModelArts uses Object Storage Service (OBS) to securely and reliably store data and models at low costs. For more details, see Object Storage Service Console Operation Gu",
"product_code":"modelarts",
"title":"What Are the Relationships Between ModelArts and Other Services?",
"uri":"modelarts_05_0003.html",
"doc_type":"usermanual",
"p_code":"221",
"code":"223"
},
{
"desc":"Log in to the console, enter the My Credentials page, and choose Access Keys > Create Access Key.In the Create Access Key dialog box that is displayed, use the login pass",
"product_code":"modelarts",
"title":"How Do I Obtain an Access Key?",
"uri":"modelarts_05_0004.html",
"doc_type":"usermanual",
"p_code":"221",
"code":"224"
},
{
"desc":"Before using ModelArts to develop AI models, data needs to be uploaded to an OBS bucket. You can log in to the OBS console to create an OBS bucket, create a folder in it,",
"product_code":"modelarts",
"title":"How Do I Upload Data to OBS?",
"uri":"modelarts_05_0013.html",
"doc_type":"usermanual",
"p_code":"221",
"code":"225"
},
{
"desc":"The AI frameworks and versions supported by ModelArts vary slightly based on the development environment notebook, training jobs, and model inference (AI application mana",
"product_code":"modelarts",
"title":"Which AI Frameworks Does ModelArts Support?",
"uri":"modelarts_05_0128.html",
"doc_type":"usermanual",
"p_code":"221",
"code":"226"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"ExeML",
"uri":"modelarts_05_0015.html",
"doc_type":"usermanual",
"p_code":"220",
"code":"227"
},
{
"desc":"ExeML is the process of automating model design, parameter tuning, and model training, compression, and deployment with the labeled data. The process is free of coding an",
"product_code":"modelarts",
"title":"What Is ExeML?",
"uri":"modelarts_05_0002.html",
"doc_type":"usermanual",
"p_code":"227",
"code":"228"
},
{
"desc":"Image classification is an image processing method that separates different classes of targets according to the features reflected in the images. With quantitative analys",
"product_code":"modelarts",
"title":"What Are Image Classification and Object Detection?",
"uri":"modelarts_05_0018.html",
"doc_type":"usermanual",
"p_code":"227",
"code":"229"
},
{
"desc":"The Train button turns to be available when the training images for an image classification project are classified into at least two categories, and each category contain",
"product_code":"modelarts",
"title":"What Should I Do When the Train Button Is Unavailable After I Create an Image Classification Project and Label the Images?",
"uri":"modelarts_05_0005.html",
"doc_type":"usermanual",
"p_code":"227",
"code":"230"
},
{
"desc":"Yes. You can add multiple labels to an image.",
"product_code":"modelarts",
"title":"Can I Add Multiple Labels to an Image for an Object Detection Project?",
"uri":"modelarts_05_0006.html",
"doc_type":"usermanual",
"p_code":"227",
"code":"231"
},
{
"desc":"Models created in ExeML are deployed as real-time services. You can add images or compile code to test the services, as well as call the APIs using the URLs.After model d",
"product_code":"modelarts",
"title":"What Type of Service Is Deployed in ExeML?",
"uri":"modelarts_05_0008.html",
"doc_type":"usermanual",
"p_code":"227",
"code":"232"
},
{
"desc":"Images in JPG, JPEG, PNG, or BMP format are supported.",
"product_code":"modelarts",
"title":"What Formats of Images Are Supported by Object Detection or Image Classification Projects?",
"uri":"modelarts_05_0010.html",
"doc_type":"usermanual",
"p_code":"227",
"code":"233"
},
{
"desc":"If an image classification or object detection algorithm of ExeML is used, after the labeled data is trained, the training result is an image error. Table 1 lists solutio",
"product_code":"modelarts",
"title":"What Do I Do If an Image Error Occurred During Model Training Using ExeML?",
"uri":"modelarts_05_0502.html",
"doc_type":"usermanual",
"p_code":"227",
"code":"234"
},
{
"desc":"ModelArts ExeML supports image classification, object detection, predictive analytics, sound classification, and text classification projects. Up to 100 ExeML projects ca",
"product_code":"modelarts",
"title":"Is There a Limit on the Number of ExeML Projects That Can Be Created?",
"uri":"modelarts_05_0179.html",
"doc_type":"usermanual",
"p_code":"227",
"code":"235"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Data Management",
"uri":"modelarts_05_0101.html",
"doc_type":"usermanual",
"p_code":"220",
"code":"236"
},
{
"desc":"Failed to use the manifest file of the published dataset to import data again.Data has been changed in the OBS directory of the published dataset, for example, images hav",
"product_code":"modelarts",
"title":"Why Does Data Fail to Be Imported Using the Manifest File?",
"uri":"modelarts_05_0103.html",
"doc_type":"usermanual",
"p_code":"236",
"code":"237"
},
{
"desc":"Images in a created dataset cannot be displayed during labeling, and they cannot be viewed by clicking them. Alternatively, the system displays a message indicating that ",
"product_code":"modelarts",
"title":"What Do I Do If Images in a Dataset Cannot Be Displayed?",
"uri":"modelarts_05_0125.html",
"doc_type":"usermanual",
"p_code":"236",
"code":"238"
},
{
"desc":"The possible cause is that the storage class of the target OBS bucket is incorrect. In this case, select a bucket of the standard storage class to import data.",
"product_code":"modelarts",
"title":"What Do I Do If Importing a Dataset Failed?",
"uri":"modelarts_05_3148.html",
"doc_type":"usermanual",
"p_code":"236",
"code":"239"
},
{
"desc":"The ModelArts console provides data visualization capabilities, which allows you to view detailed data and labeling information on the console. To learn more about the pa",
"product_code":"modelarts",
"title":"Where Are Labeling Results Stored?",
"uri":"modelarts_05_0193.html",
"doc_type":"usermanual",
"p_code":"236",
"code":"240"
},
{
"desc":"After being published, the labeling information and data in ModelArts datasets are stored as manifest files in the OBS path set for Output Dataset Path.To obtain the OBS ",
"product_code":"modelarts",
"title":"How Do I Download Labeling Results to a Local PC?",
"uri":"modelarts_05_0194.html",
"doc_type":"usermanual",
"p_code":"236",
"code":"241"
},
{
"desc":"Ensure that the created bucket and ModelArts are in the same region. Additionally, the bucket is not encrypted. ModelArts does not support encrypted OBS buckets.",
"product_code":"modelarts",
"title":"Why Is My Newly Created Bucket Unavailable?",
"uri":"modelarts_05_0509.html",
"doc_type":"usermanual",
"p_code":"236",
"code":"242"
},
{
"desc":"The version list can be zoomed in or out. Zoom out the page before searching.Click the name of the target dataset to go to the dataset overview page. Then, zoom out the V",
"product_code":"modelarts",
"title":"Why Is My New Dataset Version Unavailable in Versions?",
"uri":"modelarts_05_0511.html",
"doc_type":"usermanual",
"p_code":"236",
"code":"243"
},
{
"desc":"For data management, there are limits on the image size when you upload images to the datasets whose labeling type is object detection or image classification. The size o",
"product_code":"modelarts",
"title":"Are There Size Limits for Images to be Uploaded?",
"uri":"modelarts_05_0102.html",
"doc_type":"usermanual",
"p_code":"236",
"code":"244"
},
{
"desc":"If the issue occurs, check the video format. Only MP4 videos can be displayed and played.",
"product_code":"modelarts",
"title":"Why Cannot Videos in a Video Dataset Be Displayed or Played?",
"uri":"modelarts_05_0505.html",
"doc_type":"usermanual",
"p_code":"236",
"code":"245"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Notebook",
"uri":"modelarts_05_0067.html",
"doc_type":"usermanual",
"p_code":"220",
"code":"246"
},
{
"desc":"In a notebook instance, you can call the ModelArts MoXing API or SDK to exchange data with OBS for uploading a file to OBS or downloading a file from OBS to the notebook ",
"product_code":"modelarts",
"title":"How Do I Upload a File from a Notebook Instance to OBS or Download a File from OBS to a Notebook Instance?",
"uri":"modelarts_05_0024.html",
"doc_type":"usermanual",
"p_code":"246",
"code":"247"
},
{
"desc":"Log in to the ModelArts management console, and choose DevEnviron > Notebooks.In the notebook list, click Open in the Operation column of the target notebook instance to ",
"product_code":"modelarts",
"title":"How Do I Enable the Terminal Function in DevEnviron of ModelArts?",
"uri":"modelarts_05_0071.html",
"doc_type":"usermanual",
"p_code":"246",
"code":"248"
},
{
"desc":"If you use OBS to store the notebook instance, after you click upload, the data is directly uploaded to the target OBS path, that is, the OBS path specified when the note",
"product_code":"modelarts",
"title":"Where Will the Data Be Uploaded to?",
"uri":"modelarts_05_0045.html",
"doc_type":"usermanual",
"p_code":"246",
"code":"249"
},
{
"desc":"/cache is a temporary directory and will not be saved. After an instance using OBS storage is stopped, data in the ~work directory will be deleted. After a notebook insta",
"product_code":"modelarts",
"title":"Do Files in /cache Still Exist After a Notebook Instance is Stopped or Restarted? How Do I Avoid a Restart?",
"uri":"modelarts_05_0080.html",
"doc_type":"usermanual",
"p_code":"246",
"code":"250"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Training Jobs",
"uri":"modelarts_05_0030.html",
"doc_type":"usermanual",
"p_code":"220",
"code":"251"
},
{
"desc":"The code directory for creating a training job has limits on the size and number of files.Delete the files except the code from the code directory or save the files in ot",
"product_code":"modelarts",
"title":"What Can I Do If the Message \"Object directory size/quantity exceeds the limit\" Is Displayed When I Create a Training Job?",
"uri":"modelarts_05_0031.html",
"doc_type":"usermanual",
"p_code":"251",
"code":"252"
},
{
"desc":"When creating a training job, you can select CPU, GPU, or Ascend resources based on the size of the training job.ModelArts mounts the disk to the /cache directory. You ca",
"product_code":"modelarts",
"title":"What Are Sizes of the /cache Directories for Different Resource Specifications in the Training Environment?",
"uri":"modelarts_05_0090.html",
"doc_type":"usermanual",
"p_code":"251",
"code":"253"
},
{
"desc":"When a model references a dependency package, select a frequently-used framework to create training jobs. In addition, place the required file or installation package in ",
"product_code":"modelarts",
"title":"How Do I Create a Training Job When a Dependency Package Is Referenced in a Model?",
"uri":"modelarts_05_0063.html",
"doc_type":"usermanual",
"p_code":"251",
"code":"254"
},
{
"desc":"A third-party library may be used during job training. The following uses C++ as an example to describe how to install a third-party library.Download source code to a loc",
"product_code":"modelarts",
"title":"How Do I Install a Library That C++ Depends on?",
"uri":"modelarts_05_0088.html",
"doc_type":"usermanual",
"p_code":"251",
"code":"255"
},
{
"desc":"During job training, some parameters need to be loaded from a pre-trained model to initialize the current model. You can use the following methods to load the parameters:",
"product_code":"modelarts",
"title":"How Do I Load Some Well Trained Parameters During Job Training?",
"uri":"modelarts_05_0091.html",
"doc_type":"usermanual",
"p_code":"251",
"code":"256"
},
{
"desc":"If the training job is always queuing, the selected resources are limited in the resource pool, and the job needs to be queued. In this case, wait for resources. To speed",
"product_code":"modelarts",
"title":"Why Is a Training Job Always Queuing?",
"uri":"modelarts_05_0363.html",
"doc_type":"usermanual",
"p_code":"251",
"code":"257"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Model Management",
"uri":"modelarts_05_0016.html",
"doc_type":"usermanual",
"p_code":"220",
"code":"258"
},
{
"desc":"ModelArts does not support the import of models in .h5 format. You can convert the models in .h5 format of Keras to the TensorFlow format and then import the models to Mo",
"product_code":"modelarts",
"title":"How Do I Import the .h5 Model of Keras to ModelArts?",
"uri":"modelarts_21_0086.html",
"doc_type":"usermanual",
"p_code":"258",
"code":"259"
},
{
"desc":"ModelArts allows you to upload local models to OBS or import models stored in OBS directly into ModelArts.For details about how to import a model from OBS, see \"Importing",
"product_code":"modelarts",
"title":"How Do I Import a Model Downloaded from OBS to ModelArts?",
"uri":"modelarts_05_0124.html",
"doc_type":"usermanual",
"p_code":"258",
"code":"260"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Service Deployment",
"uri":"modelarts_05_0017.html",
"doc_type":"usermanual",
"p_code":"220",
"code":"261"
},
{
"desc":"Models can be deployed as real-time services or batch services.",
"product_code":"modelarts",
"title":"What Types of Services Can Models Be Deployed as on ModelArts?",
"uri":"modelarts_05_0012.html",
"doc_type":"usermanual",
"p_code":"261",
"code":"262"
},
{
"desc":"Before importing a model, you need to place the corresponding inference code and configuration file in the model folder. When encoding with Python, you are advised to use",
"product_code":"modelarts",
"title":"What Should I Do If a Conflict Occurs When Deploying a Model As a Real-Time Service?",
"uri":"modelarts_05_0100.html",
"doc_type":"usermanual",
"p_code":"261",
"code":"263"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Troubleshooting",
"uri":"trouble-modelarts-0000.html",
"doc_type":"usermanual",
"p_code":"",
"code":"264"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"OBS Operation Issues",
"uri":"modelarts_13_0070.html",
"doc_type":"usermanual",
"p_code":"264",
"code":"265"
},
{
"desc":"How to read the json and npy files when creating a training job.How the training job uses the cv2 library to read files.How to use the torch package in the MXNet environm",
"product_code":"modelarts",
"title":"Failed to Correctly Read Files",
"uri":"modelarts_13_0018.html",
"doc_type":"usermanual",
"p_code":"265",
"code":"266"
},
{
"desc":"After a training job is started based on TensorFlow-1.8 and the tf.gfile module is used to connect to OBS in code, the following log information is frequently printed:Thi",
"product_code":"modelarts",
"title":"Error Message Is Displayed Repeatedly When a TensorFlow-1.8 Job Is Connected to OBS",
"uri":"modelarts_13_0019.html",
"doc_type":"usermanual",
"p_code":"265",
"code":"267"
},
{
"desc":"The following error message is displayed for a ModelArts training job:The size of files to be uploaded at a time is limited to 5 GB in OBS. TensorFlow may save the summar",
"product_code":"modelarts",
"title":"TensorFlow Stops Writing TensorBoard to OBS When the Size of Written Data Reaches 5 GB",
"uri":"modelarts_13_0022.html",
"doc_type":"usermanual",
"p_code":"265",
"code":"268"
},
{
"desc":"An error occurs in the log when a model is saved in a training job. The error details are as follows:InternalError (see above for traceback): : Unable to connect to endpo",
"product_code":"modelarts",
"title":"Error \"Unable to connect to endpoint\" Error Occurs When a Model Is Saved",
"uri":"modelarts_13_0020.html",
"doc_type":"usermanual",
"p_code":"265",
"code":"269"
},
{
"desc":"When you use ModelArts, your data is stored in an OBS bucket. There is an OBS path to your data, for example, bucket_name/dir/image.jpg. ModelArts training jobs run in co",
"product_code":"modelarts",
"title":"What Do I Do If Error Message \"No such file or directory\" Is Displayed in Training Job Logs?",
"uri":"modelarts_05_0032.html",
"doc_type":"usermanual",
"p_code":"265",
"code":"270"
},
{
"desc":"The error message is displayed when MoXing is used to copy data for a training job.The possible causes are as follows:In a large-scale distributed job, multiple nodes are",
"product_code":"modelarts",
"title":"Error Message \"BrokenPipeError: Broken pipe\" Displayed When OBS Data Is Copied",
"uri":"modelarts_trouble_0042.html",
"doc_type":"usermanual",
"p_code":"265",
"code":"271"
},
{
"desc":"When TensorBoard is used to directly write data in an OBS path for a training job, an error similar to the following is displayed.The possible causes are as follows:It is",
"product_code":"modelarts",
"title":"Error Message \"ValueError: Invalid endpoint: obs.xxxx.com\" Displayed in Logs",
"uri":"modelarts_trouble_0048.html",
"doc_type":"usermanual",
"p_code":"265",
"code":"272"
},
{
"desc":"When MoXing is used to access an OBS path, the following error is displayed:ERROR:root:\nstat:404\nerrorCode:NoSuchKey\nerrorMessage:The specified key does not exist.The pos",
"product_code":"modelarts",
"title":"Error Message \"errorMessage:The specified bucket does not exist\" Displayed in Logs",
"uri":"modelarts_trouble_0035.html",
"doc_type":"usermanual",
"p_code":"265",
"code":"273"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"In-Cloud Migration Adaptation Issues",
"uri":"modelarts_13_0071.html",
"doc_type":"usermanual",
"p_code":"264",
"code":"274"
},
{
"desc":"The following error occurs in the log when a module is imported to a ModelArts training job:When a training job is imported to the module, the previous two error messages",
"product_code":"modelarts",
"title":"Failed to Import a Module",
"uri":"modelarts_13_0014.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"275"
},
{
"desc":"Perform the following operations to locate the fault:Checking Whether the Dependency Package Is AvailableChecking Whether the Dependency Package Path Can Be DetectedSumma",
"product_code":"modelarts",
"title":"Error Message \"No module named .*\" Displayed in Training Job Logs",
"uri":"modelarts_trouble_0015.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"276"
},
{
"desc":"Failed to install custom library functions for ModelArts, for example, apex.The following error occurs when a third-party package is installed in the ModelArts training e",
"product_code":"modelarts",
"title":"Failed to Install a Third-Party Package",
"uri":"modelarts_13_0015.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"277"
},
{
"desc":"The code directory fails to be downloaded during training job running, and the following error message is displayed. See Figure 1.The code directory specified during trai",
"product_code":"modelarts",
"title":"Failed to Download the Code Directory",
"uri":"modelarts_13_0023.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"278"
},
{
"desc":"The following error occurs in the ModelArts training job log:FileNotFoundError:[Errno 2]No such file or directory:'data_v.pickle'According to the error message, the file ",
"product_code":"modelarts",
"title":"Failed to Find a File in a Training Job",
"uri":"modelarts_13_0039.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"279"
},
{
"desc":"Perform the following operations to locate the fault:Checking Whether the Affected Path Is an OBS PathChecking Whether the Affected Path Is AvailableSummary and Suggestio",
"product_code":"modelarts",
"title":"Error Message \"No such file or directory\" Displayed in Training Job Logs",
"uri":"modelarts_trouble_0014.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"280"
},
{
"desc":"During the execution of a ModelArts training job, the following error message is displayed in the log and the training failed:The CUDA version of the .so file generated d",
"product_code":"modelarts",
"title":"Failed to Find the .so File During Training",
"uri":"modelarts_13_0044.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"281"
},
{
"desc":"The ModelArts training job failed to parse parameters, and the following error occurs:The parameters are not defined.In the training environment, the system may pass para",
"product_code":"modelarts",
"title":"Failed to Parse Parameters and Log Error Occurs",
"uri":"modelarts_13_0012.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"282"
},
{
"desc":"The following error message is displayed when a training job is created: Operation failed. Other running job contain train_url: /bucket-20181114/code_hxm/According to the",
"product_code":"modelarts",
"title":"Training Output Path Used by Another Job",
"uri":"modelarts_13_0029.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"283"
},
{
"desc":"When a custom image was used to create a training job of the old version, error message \"No such file or directory\" was displayed.The directory of the boot file for runni",
"product_code":"modelarts",
"title":"Failed to Find the Boot File When a Training Job Is Created Using a Custom Image",
"uri":"modelarts_13_0013.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"284"
},
{
"desc":"When a PyTorch 1.0 image is used, the following error message is displayed:\"RuntimeError: std::exception\"The possible causes are as follows:The soft link of libmkldnn in ",
"product_code":"modelarts",
"title":"Error Message \"RuntimeError: std::exception\" Displayed for a PyTorch 1.0 Engine",
"uri":"modelarts_trouble_0036.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"285"
},
{
"desc":"When MindSpore is used for training, the following error message is displayed:[ERROR] RUNTIME(3002)model execute error, retCode=0x91, [the model stream execute failed]The",
"product_code":"modelarts",
"title":"Error Message \"retCode=0x91, [the model stream execute failed]\" Displayed in MindSpore Logs",
"uri":"modelarts_trouble_0054.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"286"
},
{
"desc":"If MoXing is used to adapt to an OBS path, an error occurred when Pandas of a later version read data from an OBS file.1. 'can't decode byte xxx in position xxx'\n2. 'OSEr",
"product_code":"modelarts",
"title":"Error Occurred When Pandas Reads Data from an OBS File If MoXing Is Used to Adapt to an OBS Path",
"uri":"modelarts_trouble_0033.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"287"
},
{
"desc":"Dependency conflicts occur when other packages are installed. There are special requirements on the NumPy library. However, NumPy cannot be uninstalled. The error message",
"product_code":"modelarts",
"title":"Error Message \"Please upgrade numpy to >= xxx to use this pandas version\" Displayed in Logs",
"uri":"modelarts_trouble_0052.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"288"
},
{
"desc":"An error occurred after the engine version was reinstalled or a new CUDA package was compiled based on the existing image.1. \"RuntimeError: cuda runtime error (11) : inva",
"product_code":"modelarts",
"title":"Reinstalled CUDA Version Does Not Match the One in the Target Image",
"uri":"modelarts_trouble_0047.html",
"doc_type":"usermanual",
"p_code":"274",
"code":"289"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Memory Limit Issues",
"uri":"modelarts_13_0072.html",
"doc_type":"usermanual",
"p_code":"264",
"code":"290"
},
{
"desc":"When data, code, or a model was copied during training, the following error message was displayed.The possible causes are as follows:The disk space is insufficient.When a",
"product_code":"modelarts",
"title":"Downloading Files Timed Out or No Space Left for Reading Data",
"uri":"modelarts_13_0011.html",
"doc_type":"usermanual",
"p_code":"290",
"code":"291"
},
{
"desc":"When a ModelArts training job is running, the following error is reported in the log. As a result, data cannot be copied to the container.The container space is insuffici",
"product_code":"modelarts",
"title":"Insufficient Container Space for Copying Data",
"uri":"modelarts_13_0043.html",
"doc_type":"usermanual",
"p_code":"290",
"code":"292"
},
{
"desc":"During training job creation, error message \"No space left\" is displayed when a TensorFlow multi-node job downloads data to /cache.In a TensorFlow multi-node job, the par",
"product_code":"modelarts",
"title":"Error Message \"No space left\" Displayed When a Multi-node TensorFlow Job Downloads Data to /cache",
"uri":"modelarts_13_0025.html",
"doc_type":"usermanual",
"p_code":"290",
"code":"293"
},
{
"desc":"An error occurs during the running of a ModelArts training job, indicating that the size of the log file has reached the limit.Error information indicates that the size o",
"product_code":"modelarts",
"title":"Log File Size Reached the Upper Limit",
"uri":"modelarts_13_0032.html",
"doc_type":"usermanual",
"p_code":"290",
"code":"294"
},
{
"desc":"During the program's execution, numerous \"write line error\" messages are generated. This issue recurred each time the program ran at a specific progress.The possible caus",
"product_code":"modelarts",
"title":"Error Message \"write line error\" Displayed in Logs",
"uri":"modelarts_trouble_0031.html",
"doc_type":"usermanual",
"p_code":"290",
"code":"295"
},
{
"desc":"When data, code, or a model was copied during training, the following error message was displayed.The possible causes are as follows:The disk space is insufficient.When a",
"product_code":"modelarts",
"title":"Error Message \"No space left on device\" Displayed in Logs",
"uri":"modelarts_trouble_0041.html",
"doc_type":"usermanual",
"p_code":"290",
"code":"296"
},
{
"desc":"If a training job failed due to out of memory (OOM), possible symptoms were as follows:Error code 137 is returned.The log file contained error information with keyword ki",
"product_code":"modelarts",
"title":"Training Job Failed Due to OOM",
"uri":"modelarts_trouble_0044.html",
"doc_type":"usermanual",
"p_code":"290",
"code":"297"
},
{
"desc":"Executing a training job failed, and there is no error message in user logs. The error information is displayed in the Kubernetes job body.The possible causes are as foll",
"product_code":"modelarts",
"title":"Error Message \"Pod The node was low on resource:[DiskPressure]\" Displayed in the Kubernetes Job Body",
"uri":"modelarts_trouble_0040.html",
"doc_type":"usermanual",
"p_code":"290",
"code":"298"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Internet Access Issues",
"uri":"modelarts_13_0077.html",
"doc_type":"usermanual",
"p_code":"264",
"code":"299"
},
{
"desc":"When PyTorch is used, the following error message is displayed in logs after pretrained in torchvision.models is set to True:'OSError: [Errno 101] Network is unreachable'",
"product_code":"modelarts",
"title":"Error Message \"Network is unreachable\" Displayed in Logs",
"uri":"modelarts_trouble_0034.html",
"doc_type":"usermanual",
"p_code":"299",
"code":"300"
},
{
"desc":"In a running training job, a URL connection timeout error occurs.For security purposes, ModelArts is not allowed to access the Internet to download data.Download the requ",
"product_code":"modelarts",
"title":"URL Connection Timed Out in a Running Training Job",
"uri":"modelarts_13_0021.html",
"doc_type":"usermanual",
"p_code":"299",
"code":"301"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Permission Issues",
"uri":"modelarts_13_0078.html",
"doc_type":"usermanual",
"p_code":"264",
"code":"302"
},
{
"desc":"When a training job accessed OBS, an error occurred.The possible causes are as follows (see Python > Troubleshooting > OBS Server-Side Error Codes in Object Storage Servi",
"product_code":"modelarts",
"title":"Error Message \"reason:Forbidden\" Displayed in Logs",
"uri":"modelarts_trouble_0045.html",
"doc_type":"usermanual",
"p_code":"302",
"code":"303"
},
{
"desc":"When a training job accesses the attached EFS disks or executes the .sh boot script, an error occurs.[Errno 13]Permission denied: '/xxx/xxxx'Error logbash: /bin/ln: Permi",
"product_code":"modelarts",
"title":"Error Message \"Permission denied\" Displayed in Logs",
"uri":"modelarts_trouble_0046.html",
"doc_type":"usermanual",
"p_code":"302",
"code":"304"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"GPU Issues",
"uri":"modelarts_13_0079.html",
"doc_type":"usermanual",
"p_code":"264",
"code":"305"
},
{
"desc":"An error similar to the following occurs during the running of the program:1. 'failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected'\n2. 'No CU",
"product_code":"modelarts",
"title":"Error Message \"No CUDA-capable device is detected\" Displayed in Logs",
"uri":"modelarts_trouble_0032.html",
"doc_type":"usermanual",
"p_code":"305",
"code":"306"
},
{
"desc":"When PyTorch was used for distributed training, the following error occurred.The possible causes are as follows:If data had been copied before this issue occurred, data r",
"product_code":"modelarts",
"title":"Error Message \"RuntimeError: connect() timed out\" Displayed in Logs",
"uri":"modelarts_trouble_0043.html",
"doc_type":"usermanual",
"p_code":"305",
"code":"307"
},
{
"desc":"A training job failed, and the following error was printed in logs.The possible causes are as follows:The CUDA_VISIBLE_DEVICES setting does not comply with job specificat",
"product_code":"modelarts",
"title":"Error Message \"cuda runtime error (10) : invalid device ordinal at xxx\" Displayed in Logs",
"uri":"modelarts_trouble_0049.html",
"doc_type":"usermanual",
"p_code":"305",
"code":"308"
},
{
"desc":"When PyTorch was used to start multiple processes, the following error message was displayed:RuntimeError: Cannot re-initialize CUDA in forked subprocessThe possible caus",
"product_code":"modelarts",
"title":"Error Message \"RuntimeError: Cannot re-initialize CUDA in forked subprocess\" Displayed in Logs",
"uri":"modelarts_trouble_0051.html",
"doc_type":"usermanual",
"p_code":"305",
"code":"309"
},
{
"desc":"The following error message is displayed during the running of a ModelArts training job:According to error information, the error cause is that the training job running p",
"product_code":"modelarts",
"title":"No GPU Detected in a Training Job",
"uri":"modelarts_13_0033.html",
"doc_type":"usermanual",
"p_code":"305",
"code":"310"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Service Code Issues",
"uri":"modelarts_13_0073.html",
"doc_type":"usermanual",
"p_code":"264",
"code":"311"
},
{
"desc":"When Pandas was used to read CSV data, the following error was printed in logs, and the training job failed:pandas.errors.ParserError: Error tokenizing data. C error: Exp",
"product_code":"modelarts",
"title":"Error Message \"pandas.errors.ParserError: Error tokenizing data. C error: Expected .* fields\" Displayed in Logs",
"uri":"modelarts_trouble_0050.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"312"
},
{
"desc":"After PyTorch 1.3 was upgraded to 1.4, the following error message was displayed:\"RuntimeError:max_pool2d_with_indices_out_cuda_frame failed with error code 0\"The possibl",
"product_code":"modelarts",
"title":"Error Message \"max_pool2d_with_indices_out_cuda_frame failed with error code 0\" Displayed in Logs",
"uri":"modelarts_trouble_0037.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"313"
},
{
"desc":"The training job failed, and error code 139 is returned.The possible causes are as follows:Certain pip packages in the pip source have been updated, leading to data incom",
"product_code":"modelarts",
"title":"Training Job Failed with Error Code 139",
"uri":"modelarts_trouble_0039.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"314"
},
{
"desc":"Before creating a training job, use the ModelArts development environment to debug the training code to maximally eliminate errors in code migration.",
"product_code":"modelarts",
"title":"Debugging Training Code in the Cloud Environment If a Training Job Failed",
"uri":"modelarts_trouble_0057.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"315"
},
{
"desc":"The following error message was displayed during training:TypeError: '(slice(0, 13184, None), slice(None, None, None))' is an invalid keyThe possible causes are as follow",
"product_code":"modelarts",
"title":"Error Message \"'(slice(0, 13184, None), slice(None, None, None))' is an invalid key\" Displayed in Logs",
"uri":"modelarts_trouble_0059.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"316"
},
{
"desc":"The following error message was displayed during training:DataFrame.dtypes for data must be int, float or boolThe possible causes are as follows:The training data is not ",
"product_code":"modelarts",
"title":"Error Message \"DataFrame.dtypes for data must be int, float or bool\" Displayed in Logs",
"uri":"modelarts_trouble_0058.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"317"
},
{
"desc":"The following error message was displayed during PyTorch training:RuntimeError: cuDNN error: CUDNN_STATUS_NOT_SUPPORTED. This error may appear if you passed in a non-cont",
"product_code":"modelarts",
"title":"Error Message \"CUDNN_STATUS_NOT_SUPPORTED.\" Is Printed in Logs",
"uri":"modelarts_trouble_0056.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"318"
},
{
"desc":"When pandas.to_datetime was used to convert time, the following error message was displayed:pandas._libs.tslibs.np_datetime.OutOfBoundsDatetime: Out of bounds nanosecond ",
"product_code":"modelarts",
"title":"Error Message \"Out of bounds nanosecond timestamp\" Displayed in Logs",
"uri":"modelarts_trouble_0053.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"319"
},
{
"desc":"After Keras was upgraded to 2.3.0 or later, the following error message was displayed:TypeError: Unexpected keyword argument passed to optimizer: learning_rateThe possibl",
"product_code":"modelarts",
"title":"Error Message \"Unexpected keyword argument passed to optimizer\" Displayed in Logs",
"uri":"modelarts_trouble_0055.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"320"
},
{
"desc":"An NCCL debug log level is set in a distributed job executed using a PyTorch image.import os\nos.environ[\"NCCL_DEBUG\"] = \"INFO\"The following error message was displayed.Th",
"product_code":"modelarts",
"title":"Error Message \"no socket interface found\" Displayed in Logs",
"uri":"modelarts_trouble_0038.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"321"
},
{
"desc":"The following error occurs when tf.variable is used across multiple machines and multiple GPUs: WARNING:tensorflow:Gradient is None for variable:v0/tower_0/UNET_v7/sub_pi",
"product_code":"modelarts",
"title":"tf.variable Unavailable for Distributed TensorFlow",
"uri":"modelarts_13_0016.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"322"
},
{
"desc":"When kv_store = mxnet.kv.create('dist_async') is used to create kvstore, the program is blocked. For example, run the following code. If end is not displayed, the program",
"product_code":"modelarts",
"title":"Creating a KVStore Using MXNet Blocked and No Error Reported",
"uri":"modelarts_13_0026.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"323"
},
{
"desc":"The following error occurs during the running of the training job log: RuntimeError: CUDA error: uncorrectable ECC error encounteredIf a job fails to be executed due to a",
"product_code":"modelarts",
"title":"Training Job Failed Due to an ECC Error",
"uri":"modelarts_13_0028.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"324"
},
{
"desc":"An error occurs for a ModelArts training job.The training failed because the recursion depth exceeded the default recursion depth of Python.If the maximum recursion depth",
"product_code":"modelarts",
"title":"Training Job Failed Because the Maximum Recursion Depth Is Exceeded",
"uri":"modelarts_13_0034.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"325"
},
{
"desc":"When a training job is created using a built-in algorithm, the training failed with the following error message in the log:Non-rectangles are used for labeling training s",
"product_code":"modelarts",
"title":"Training Using a Built-in Algorithm Failed Due to a bndbox Error",
"uri":"modelarts_13_0041.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"326"
},
{
"desc":"When Algorithm Source is set to Custom during training job creation, the training job status is Reviewing Job Initialization.When a custom image is running for the first ",
"product_code":"modelarts",
"title":"Training Job in \"Reviewing Job Initialization\" State",
"uri":"modelarts_13_0030.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"327"
},
{
"desc":"The training fails and the following error information is displayed in the log.According to the log, the exit code of the training job is 137. The training process starts",
"product_code":"modelarts",
"title":"Training Job Process Exits Unexpectedly",
"uri":"modelarts_13_0074.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"328"
},
{
"desc":"The training job process is stopped and the logs are interrupted.CPU soft lockThe decompression of a large number of files may cause CPU soft lock and node restart. You c",
"product_code":"modelarts",
"title":"Stopped Training Job Process",
"uri":"modelarts_13_0075.html",
"doc_type":"usermanual",
"p_code":"311",
"code":"329"
},
{
"desc":"HUAWEI CLOUD Help Center presents technical documents to help you quickly get started with HUAWEI CLOUD services. The technical documents include Service Overview, Price Details, Purchase Guide, User Guide, API Reference, Best Practices, FAQs, and Videos.",
"product_code":"modelarts",
"title":"Change History",
"uri":"modelarts_04_0099.html",
"doc_type":"usermanual",
"p_code":"",
"code":"330"
}
]
View Git Blame Copy Permalink