PanguLargeModels
PanguLargeModels
All results for "
" in this service
All results for "
" in this service
Service Overview
What Is PanguLM?
Product Advantages
Application Scenarios
Application Scenarios of Large Models
Application Scenarios of Agents
Functions
Workspace Management
Data Engineering
Model Development
Agent Development
Model Capabilities and Specifications
Third-Party Large Models
Basic Knowledge
Basic Process of Large Model Development
Basic Concepts
Security
Shared Responsibilities
User Authentication and Access Control
Data Protection
Auditing
Permissions Management
Notes and Constraints
Related Services
Billing
Billing Overview
Billing Mode
Billing Item
Changing the Billing Mode
Renewal
Arrears
Billing Termination
Billing FAQ
What Are the Differences Between Yearly/Monthly and Pay-per-Use Billing?
Which Is More Cost-Effective, Yearly/Monthly or Pay-per-Use Billing?
Can a Resource Be Billed Using Both Yearly/Monthly and Pay-per-Use Modes?
Can I Switch Between Yearly/Monthly and Pay-per-Use Billing Modes?
How Do I Renew Resources?
Getting Started
Using the Pangu Pre-trained NLP Model for Text Dialog
Using the Pangu NLP Model to Create a Python Coding Assistant Application
User Guide
Process of Using PanguLM
Preparations
Applying for Trial Use of ModelArts Studio Large Model Development Platform
Subscribing to the PanguLM Service
Configuring Service Access Authorization
Creating and Managing Workspaces
Workspace Overview
Creating and Managing Workspaces
Managing Workspace Members
Accessing Models in the Model Square
Using Data Engineering to Create a Dataset
Introduction to Data Engineering
Process of Using Data Engineering
Dataset Format Requirements
Format Requirements for Text Datasets
Format Requirements for Other Datasets
Importing Data to the Pangu Platform
Processing Datasets
Dataset Processing Scenarios
Processing Text Datasets
Processing Text Datasets
Synthesizing Text Datasets
Labeling Text Datasets
Combining Text Datasets Based on a Specific Ratio
Processing Other Datasets
Managing Processing Operators
Introduction to Preset Processing Operators
Text Dataset Processing Operators
Custom Data Processing Operators
Operator Configuration File Specifications
Operator Package Development Specifications
Typical Operator Development Examples
Managing Custom Operators
Managing Processed Datasets
Managing Processing Task Resources
Using a Processing Template
Managing Processing Models
Generating a Dataset in a Processing Task
Publishing a Dataset
Dataset Publishing Scenarios
Publishing Text Datasets
Evaluating Text Datasets
Publishing Text Datasets
Publishing Other Datasets
Managing Published Datasets
Converting the Dataset Format
Constraints
Format Conversion Process
Conversion Operators
Common Errors and Solutions for Data Engineering
Developing a DeepSeek Model
DeepSeek Models
Using Data Engineering to Build a DeepSeek Model Dataset
Deploying a DeepSeek Model
Evaluating a DeepSeek Model
Creating a DeepSeek Model Evaluation Dataset
Creating an API Service
Creating a DeepSeek Model Evaluation Job
Viewing the DeepSeek Model Evaluation Report
Managing DeepSeek Model Evaluation Jobs
Calling a DeepSeek Model
Developing a Third-Party Model
Using Data Engineering to Build a Third-Party Model Dataset
Deploying a Third-Party Model
Model Deployment Modes and Description
Creating a Third-Party Model Deployment Task
Viewing Details About a Third-Party Model Deployment Task
Managing Third-Party Model Deployment Tasks
Evaluating a Third-Party Model
Creating a Third-Party Model Evaluation Dataset
Creating an API Service
Creating a Third-Party Model Evaluation Job
Viewing the Third-Party Model Evaluation Report
Managing Third-Party Model Training Jobs
Calling a Third-Party Model
Using the Experience Center Function to Call a Third-Party Model
Using APIs to Call a Third-Party Model
Collecting Third-PartyModel Call Statistics
Third-Party Model Training and Reference High Availability
Training Log Failure Analysis
Developing a Prompt Engineering Project
What Is Prompt Engineering?
Obtaining a Prompt Template
Writing Prompts
Creating a Prompt Engineering Project
Writing Prompts
Previewing Prompt Outcomes
Comparing Outcomes Between Prompts
Setting Candidate Prompts
Comparing Outcomes Between Prompts
Evaluating the Prompt Outcomes in Batches
Creating a Prompt Evaluation Dataset
Creating a Prompt Evaluation Task
Viewing the Prompt Evaluation Result
Publishing Prompts
Developing a Pangu Domain-specific Application
Introduction to Industrial Application Orchestration
Orchestrating Industrial Applications
Creating Application Components
Functions of Preset Components
Creating a Script Component
Creating an Algorithm Package Component
Deploying the Algorithm Package Component
Creating a Static Application
Deploying a Static Application
Calling a Static Application
Industrial Application Orchestration Practices
Example: Using Preset Components to Create a Containerized Data Access Application
Developing an Agent
Agent Development Platform Overview
Platform Overview
Procedure
Using a Preset Agent in the Application Library
Quickly Setting Up an Agent Application
Developing a Single-Agent Application
Basic Settings
Configuring Prompts
Configuring the Agent Scheduling Mode
Adding Skills to an Application
Configuring a Plug-in
Configuring a Workflow
Configuring a Knowledge Base
Configuring an MCP Service
Improving Dialog Experience of Applications
Debugging and Publishing an Application
Debugging an Application
Publishing an Application as an API Service
Using APIs to Call an Application
Developing a Workflow Application
Workflow Introduction
Dialogue-based Workflows and Task-based Workflows
Creating a Workflow
Configuring Chat Memory
Configuring a Multi-Agent Application Workflow
Configuring a Multi-Agent Application Workflow
Debugging and Publishing a Workflow
Using APIs to Call a Workflow
Workflow Node Configuration Reference
Start and End Nodes
LLM Node
Knowledge Repo Node
IntentDetection Node
Plugin Node
Branch Node
Code Node
Message Node
Questioner Node
Loop Node
Variable Assignment Node
Aggregation Node
Input Node
Workflow Node
MCP Service Node
Agent Node
Managing Workflows
Managing Agent Platform Plug-ins
Managing Plug-ins
Introduction to Plug-ins
Creating a Plug-in
Creating a Plug-in Based on an API
Importing Plug-ins Using JSON Files
Managing Plug-ins
Managing Knowledge Bases
Knowledge Base Introduction
Creating a Knowledge Base and Uploading Documents
Knowledge Base Hit Test
Managing Knowledge Bases
Managing MCP Services
MCP Service Introduction
Creating an MCP Service
Managing MCP Services
Subscribing to the MCP Service
Common Errors and Solutions During Agent Development
Managing Agents
Managing Workspace Assets
Introduction to Pangu Model Workspace Assets
Managing Pangu Data Assets
Managing Pangu Model Assets
Managing Pangu Component Assets
Managing Resource Pools
Creating an Edge Resource Pool
Best Practices
Prompt Writing Practices
General Tips for Prompt Writing
Advanced Approaches for Prompt Writing
Setting the Context and Persona
Understanding Task Logic
Chain-of-Thought Prompting
Analyzing the Model's Reasoning Logic
Prompt Application Examples
Using Prompts to Implement Intent Alignment in an Intelligent Customer Service System
Using Prompts to Generate Interview Questions
Practice of Building a Dataset
Building an Incremental Pre-training Dataset for the NLP Model
Obtaining Source Data
Preprocessing Data
Importing Data
Processing Datasets
Evaluating Datasets
Combining and Publishing Datasets
Building a Fine-Tuning Dataset for the NLP Model
Obtaining Source Data
Preprocessing Data
Importing Data
Processing Datasets
Evaluating Datasets
Combining and Publishing Datasets
Agent Application Practices
Building AI Research Assistants Without Coding
Solution Design
Build Process
Creating an Application
Typical Problems
Building an Intelligent Assistant Workflow with Low Code
Solution Design
Build Process
API Reference
Before You Start
Overview
API Calling
Request URI
Concepts
Calling REST APIs
Making an API Request
Authentication
Response
API
Model Inference APIs
Third-Party Models
Third-Party NLP Models
Qwen Third-Party VL Model
Data Engineering APIs
Querying Data Lineages
Permanently Deleting a Dataset
Agent APIs
Calling an Application
Calling a Workflow
Token Calculator
Appendix
Status Codes
Error Codes
Obtaining the Project ID
Obtaining the Model Deployment ID
FAQs
FAQs
FAQs Related to LLM Concepts
How Do I Evaluate and Protect the Safety of Pangu Models?
How Can an LLM Be Effectively Trained to Adapt to Intelligent Customer Service Scenarios?
FAQs Related to Permissions
Why Cannot I Find a Workspace on ModelArts Studio?
What Permissions Are Required for an IAM Account to Use the ModelArts Studio?
FAQs Related to Data Operations
Common Errors and Solutions for Data Import Tasks
Failed to Parse the Import Task File
Insufficient Resources for the Import Task
No Permission Is Displayed on the Import Task Page
User Has Not Subscribed to OBS-Related Services
Why Cannot I Select a Single File from OBS for Upload During Data Import?
How Do I Upload Local Data to ModelArts Studio?
Common Errors and Solutions for Data Processing Tasks
Processing Task Failure Caused by Task Schedule Error
Created Datasets Cannot Be Found During Data Processing Task Creation
Where Can I Find the Processed Dataset?
Synthesis Task Failure with an Error Message Indicating Task Execution Failure
FAQs Related to LLM Fine-Tuning and Training
How Do I Enable Models to Learn Unsupervised Domain-Specific Knowledge If the Data Volume Is Insufficient for Incremental Pre-training?
How Do I Adjust Training Parameters to Maximize the Pangu Model Performance?
How Do I Determine Whether the Pangu Model Training Status Is Normal?
How Do I Evaluate Whether the Fine-Tuned Pangu Model Is Normal?
How Do I Adjust Inference Parameters to Maximize the Pangu Model Performance?
Why Does the Fine-Tuned Pangu Model Always Repeat the Same Answer?
Why Does the Fine-Tuned Pangu Model Generate Garbled Characters?
Why Is the Answer of the Fine-Tuned Pangu Model Truncated Abnormally?
Why Can the Fine-Tuned Pangu Model Only Answer the Questions in the Training Sample?
Why Does the Fine-Tuned Pangu Model Return Different Answers to the Same Question in the Training Sample?
Why Is the Performance of the Fine-Tuned Pangu Model in Actual Scenarios Worse Than That During Evaluation?
Why Is the Performance of the Fine-Tuned Pangu Model Unsatisfactory in Multi-Turn Dialogues?
Why Is the Performance of the Fine-Tuned Pangu Model Unsatisfactory When the Data Volume Is Sufficient?
Why Is the Performance of the Fine-Tuned Pangu Model Unsatisfactory Even Though Both the Data Volume and Quality Meet Requirements?
FAQs Related to Model Deployment
How Do I Obtain the Model Deployment ID?
FAQs Related to LLM Usage
Can the Persona of a Pangu Model Be Customized?
How Do I View Historical Versions of a Preset Model?
What Is the Mapping Between a Training or Inference Unit and Computing Power?
FAQs Related to Prompt Engineering
How Do I Improve the Accuracy of an LLM in Complex Inference Tasks Using Prompts
How Do I Ask the Model to Respond in a Specified Style or Format?
How Do I Analyze the Root Cause of Incorrect Outputs of a Foundation Model?
Why Do Prompts That Work Well on Other Models Not Effective on Pangu Models?
How Do I Determine Whether to Adjust Prompts or Use Scenario-Specific Fine-Tuning?
General Reference
Glossary
Service Level Agreement
White Papers
Endpoints
Permissions