You can also dial

‭(0571) 2808 6265

Caicloud Clever
Artificial Intelligence Cloud Platform

Details

Product Advantage

One-stop

Integrate data import, data processing, model development, model training, model evaluation and service online to provide a one-stop comprehensive deep learning modeling process to quickly build AI services.

Efficiency

Efficiently manage and allocate hardware resources, schedule model training and model services to bring a speedy experience.

Ease of Use

Simple operation and quick to start with minimal requirement of user operation, maintenance, software and hardware configuration.

Support Multiple Frames

Support multiple deep learning and machine learning frameworks such as TensorFlow, PyTorch, Caffe and MXNet.

Product Features

clever icon(英) clever icon(英)

AI-DevOps

Provides one-stop multi-dimensional AI-DevOps implementation to support data and models, code hosting, model development, training, and real-time services.

clever icon(英) clever icon(英)

Database

Supports versioned data set management, linked platform development, training and pipeline. Easy to click & use.

clever icon(英) clever icon(英)

Model Repos

Supports versioned model set management, multiple model frameworks, and one-click deployment services.

clever icon(英) clever icon(英)

Code Management

Pre-built-in GitLab code hosting service to support code Git management and training code traceability.

Data Source Management

Supports four types of data source management: MySQL, FTP, HDFS, and Hive, and supports offline batch data synchronization.

clever icon(英) clever icon(英)

Model Training

Supports TensorFlow, Keras, PyTorch, Caffe, MXNet frameworks; supports GPU training acceleration, and multi-card distributed training on multiple machines.

Model Optimization

One-click model optimization, shorten inference time, and accelerate model training.

clever icon(英) clever icon(英)

Algorithm Solutions

Supports landing AI solutions by directly copying them.

clever icon(英) clever icon(英)

AI Pipelines

Supports AI customized pipeline and tasks such as data cleaning, model training, evaluation, and service.

clever icon(英) clever icon(英)

Model Services

Supports one-click deployment model service, GPU sharing, A / B test, and customized container image.

clever icon(英) clever icon(英)

GPU Sharing

Supports GPU multi-service sharing.

clever icon(英) clever icon(英)

Model Evaluation

Supports model evaluation of preset and customized rules and multi-model comparison.

Multi-tenant Management

Fits the enterprise multi-tenant scenario and supports resource isolation between tenants.

RBAC

Supports fine-grained authentication permissions.

Demo

展示产品
展示产品

展示产品

Component Library

展示产品

Working Space

展示产品

Modal Service

展示产品

Application Scenario

clever

Data Preprocessing

Support object-oriented storage, online data cleaning, online data labelling system, raw data pre-processing with one-click to import into platform storage.

clever

Management of Training Jobs

Support GPU accelerated training, support distributed training, and visually deploy training jobs. Use priority queues to create batch training tasks.

clever

Rapid Model Iteration

Given agile model development, training results are imported into the model warehouse thus can be released and deployed. Support online model version management, model CI pipelines, model service grayscale release and elastic scaling.

Product Update

2017-07-13

Supports online TensorFlow training

  • Supports online TensorFlow training
  • Supports multi-training task management
  • Supports real-time monitoring of task information
  • Supports online code editing
  • Supports GUI file management

2018-08-18

Supports distributed TensorFlow training

  • Supports distributed TensorFlow training
  • Supports multi-machine multi-card GPU accelerated training
  • Supports cloud shared storage
  • Supports analyzing model Signature
  • Supports online image labelling

2018-11-08

Supports multiple WorkSpaces

  • Supports multiple WorkSpaces
  • Supports PyTorch training
  • Supports PyTorch model services
  • Supports model development in customized environment
  • Supports model training in customized environment

2019-01-16

Supports AI-DevOps

  • Supports AI-DevOps
  • Supports engineering AI project management
  • Supports visualized AI process orchestration
  • Supports versionized data set management
  • Supports versionized model set management

2019-01-31

Supports Spark data processing

  • Supports Spark data processing
  • Supports multi-repo integration
  • Supports Keras training
  • Supports Caffe training
  • Supports scikit-learn training

2019-07-04

Supports full-featured online development work space

  • Supports full-featured online development work space
  • Supports Git code management
  • Supports Spark tasks
  • Supports model A / B testing
  • Supports model evaluation

2020-03-11

Supports Auto ML - automatic optimization of hyperparameter tuning

  • Supports Auto ML - automatic optimization of hyperparameter tuning
  • Supports training models using preset algorithms
  • Supports Multi-Jupyter
  • Model service GPU sharing
  • Optimizes data labeling

2020-07-13

Supports landing AI solutions by directly copying them

  • Supports landing AI solutions by directly copying them
  • Supports traditional machine learning online inference
  • Supports model training visualization
  • Model training resource quota recommendation
  • AI pipeline optimization

2020-11-11

Supports data source management, offline data synchronization

  • Supports data source management, offline data synchronization
  • Built-in multiple evaluation indicators
  • Supports ONNX and Caffe model optimization
  • Supports vscode remote connection to Jupyter
  • Comprehensive display of the operation of platform services and resources

Help Customers Succeed

compass
compass
compass
compass

Contact Us Now

To make sure to provide more accurate customized solutions, please fill in this pre-service form with real information. You can also dial ‭(0571) 2808 6265

Thank you,

Your message has been received

OK