Tutorials

• 1. Introduction
  • 1.1. Prerequisites
  • 1.2. Running these tutorials
• 2. PyTorch
  • 2.1. Introduction to PopTorch - running a simple model
  • 2.2. Efficient data loading with PopTorch
  • 2.3. Tutorial on BERT Fine-tuning on IPU
  • 2.4. Half and mixed precision in PopTorch
  • 2.5. Observing tensors in PopTorch
  • 2.6. PopTorch Parallel Execution Using Pipelining
  • 2.7. Training a Hugging Face model on the IPU using a local dataset
• 3. PopXL
  • 3.1. PopXL and popxl.addons
  • 3.2. PopXL Custom Optimiser
  • 3.3. Data parallelism
  • 3.4. Pipelining
  • 3.5. Remote variables and RTS
  • 3.6. Phased Execution
• 4. Poplar
  • 4.1. Tutorial 1: programs and variables
  • 4.2. Tutorial 2: using PopLibs
  • 4.3. Tutorial 3: writing vertex code
  • 4.4. Tutorial 4: profiling output
  • 4.5. Tutorial 5: matrix-vector multiplication
  • 4.6. Tutorial 6: matrix-vector multiplication optimisation
• 5. TensorFlow 1
  • 5.1. Introduction to TensorFlow 1 on the IPU
  • 5.2. TensorFlow 1 on the IPU: training a model using half- and mixed-precision
  • 5.3. TensorFlow 1 Pipelining
• 6. TensorFlow 2
  • 6.1. Using Infeed and Outfeed Queues in TensorFlow 2
  • 6.2. Keras tutorial: How to run on IPU
  • 6.3. Using TensorBoard in TensorFlow 2 on the IPU
• 7. PopVision
  • 7.1. Tutorial: Accessing profiling information
  • 7.2. Tutorial: Lightweight profiling
  • 7.3. Tutorial: Reading PVTI files with libpva
  • 7.4. Tutorial: Instrumenting applications
• 8. Standard Tools
  • 8.1. Using IPUs from Jupyter Notebooks
  • 8.2. Using VS Code with the Poplar SDK and IPUs
• 9. Next steps
• 10. Trademarks & copyright