A quick Example

• Tutorial 1: deeptb-sk baseline model
  • Introduction
    • Version Features
    • Learning Objectives
  • 1. Calculating Band Structure for a Given Structure
  • 2. Extracting SK Parameter Files for a Given System
• Tutorial 2: data preparation for deeptb-sk model
  • Introduction
    • Version Features
    • Learning Objectives
  • 1. Prepare Data for DeePTB-SK Model
    • Pre-requisites
    • 1.1 ABACUS Case:
    • Band structure analysis
    • Finally, the training data set format for DeePTB-SK model is obtained.
    • 1.2. VASP Case:
    • Finally, the training data set format for DeePTB-SK model is obtained.
  • 2. Training DeePTB-SK Model using vasp data
    • 2.1 Extract initial empirical SK parameters
    • 2.1 model training (briefly introduce)
• Tutorial 3: Training deeptb-sk model for Silicon
  • Introduction
    • Version Features
    • Learning Objectives
  • Method Practice
    • 1. data preparation
    • **2. Model Training **
      • 2.1 Extract Initial Experience sktb Model
      • 2.2 Generate Training Input Control Parameters
      • 2.3.1 Training the DeePTB-SK model for perfect lattice
      • 2.3 Training on MD data (bond length dependence)
      • 2.4 Training Environment Correction
• Tutorial 4: Training deeptb-e3 model for Silicon
  • Introduction
    • Version Features
    • Learning Objectives
  • Content
  • Method:
    • What is DeePTB-E3:
    • What to do with DeePTB-E3
  • Practice
    • 1. Prepare the Data
    • 2. Model Training
  • More Method
    • Parametrization of the Equivariant Representation:
    • Invariant Representation of Overlap Integrals
    • Model Architecture