检测到您已登录华为云国际站账号，为了您更好的体验，建议您访问国际站服务网站 https://www.huaweicloud.com/intl/zh-cn
不再显示此消息
检测到您已登录华为云国际站账号，为了您更好的体验，建议您访问国际站服务网站 https://www.huaweicloud.com/intl/zh-cn
不再显示此消息
Exporting Data from ModelArts to OBS You can select certain data or obtain the required data by setting the filtering criteria. The data can be saved in OBS for future export. In this way, you can create a dataset. The historical export records can be viewed in task history.
You are advised to use the following local cache method: import moxing.tensorflow as mox mox.cache() Parent topic: OBS Operation Issues
Table 2 Data source types Type Example OBS "source":"s3://path-to-jpg" Content "source":"content://I love machine learning" Table 3 annotation objects Parameter Mandatory Description type Yes Label type.
For details about OBS error codes, see OBS Server-Side Error Codes.
Table 2 Parameters of the meta model source Parameter Description Meta Model OBS path for storing the meta model. The OBS path cannot contain spaces. Otherwise, the AI application fails to be created.
Figure 1 Importing a meta model from OBS Table 2 Meta model source parameters Parameter Description Meta Model Source Select OBS. Meta Model OBS path for storing the meta model. The OBS path cannot contain spaces. Otherwise, the model creation will fail.
How Do I Improve Training Efficiency While Reducing Interaction with OBS in ModelArts? Scenarios When you use ModelArts for custom deep learning training, training data is typically stored in OBS.
How Do I Use the pandas Library to Process Data in OBS Buckets on a ModelArts Notebook Instance? Download data from OBS to a notebook instance. For details, see Downloading a File from JupyterLab to a Local Path.
Notebook instances with OBS storage mounted can synchronize files from OBS to JupyterLab using the JupyterLab upload and download functions. The files on the Terminal page are the same as those in JupyterLab.
Parent topic: Importing Data from OBS
Parent topic: OBS Operation Issues
Failed to Download a pip Package When a Model Is Created Using OBS Symptom Creating a model using OBS failed. Logs showed that downloading the pip package failed, for example, downloading the NumPy 1.16 package failed.
Error "stat:403 reason:Forbidden" Is Displayed in Logs When a Training Job Accesses OBS Symptom When a training job accesses OBS, an error occurs.
Table 2 Data source types Type Example OBS "source":"s3://path-to-jpg" Content "source":"content://I love machine learning" Table 3 annotation objects Parameter Mandatory Description type Yes Label type.
Possible Causes When a model is imported through OBS, ModelArts copies all files and folders in the specified OBS directory to a path specified in the image. You can obtain the path in the image by using self.model_path.
Parent topic: OBS Management
trained model or other training script data is stored. obs_path: OBS path.
Using the SDK to Debug a Multi-Node Distributed Training Job Replace the OBS paths in the debugging code with your OBS paths. PyTorch is used to write debugging code in this document. The process is the same for different AI frameworks.
You can obtain the OBS path set for Output Dataset Path. Log in to the OBS management console and locate the version directory from the obtained OBS path to obtain the labeling result of the dataset.
Parsing a Manifest File Parse a manifest file in either a local or OBS path.
