nnc
a deep learning framework from ccv
  • Overview
    • What’s NNC?
    • 1. Tensors, Commands and Streams
    • 2. Computation Graph
    • 3. Symbolic Graph
    • 4. Dynamic Graph
    • 5. Common Neural Network Primitives
    • Supplementary Materials
      • Toll-Free Bridging
      • Automatic Differentiation
      • while Type Sub-Graph
      • case..of Type Sub-Graph
      • Limits and Constraints
  • Technicals
    • The NNC Tensor Allocation Algorithm
      • Tensor Representation
      • Loop Representation
      • The Problem Definition
      • The Core Algorithm
      • Basic Structure
      • Candidate Selection
      • Insertion
      • Intuition
      • Loop
      • Multi-view Tensor
      • Loop with Efficient Tensor Allocation
      • Sub-Computation Graph
      • Conclusion
    • NNC Static Schedule A Graph
      • Stream
      • Static Schedule
      • while and case..of
    • NNC Dynamic Graph Execution
      • Naming The Variable
      • Tracing The Operation
      • Optimizations
      • Interoperability
      • Future Optimizations
      • Some Maybes
    • NNC Common Neural Network Primitives
      • Model
      • Model IO
      • Fit, Evaluate, Backward, Apply Gradients
    • NNC Dataframe
      • Operations on the Data
      • Iteration
      • Map
      • Reduce
      • Others
      • Use Dataframe with Addons
    • NNC Micro Ops
      • Describe Micro Ops
      • Parameters
      • Simplification
      • Loop-Fusion
      • Variable Substitution
      • Automatic Differentiation
  • API Reference
    • Level 0 API
      • Essentials
    • Level 1 API
      • The concept of meta-ops in Jittor is amazing
      • Tensors
      • Commands
      • Streams
      • Micro Ops
    • Level 2 API
      • Essentials
      • Others
    • Level 3 API
      • Essentials
      • Others
    • Level 3.5 API
      • Construct “switch” control structure in symbolic graph
      • Symbolic graph simplification
      • Construct a “while” loop in a symbolic graph
      • Automatic Differentiation
      • While Loop Essentials
      • Branching
      • Gradient-based Optimization
      • Graph Simplification
      • Automatic Graph Parallelization
      • Memory Compression
      • While Loop Others
    • Level 4 API
      • Essentials
    • Level 5 API
      • What is “dataframe” in ML?
      • Why to support comma-separated-values files in dataframe?
      • Models, layers, and Keras
      • Dataframe API
      • Dataframe Add-ons
      • Dataframe CSV Support
      • Model API
      • Model Add-ons
    • Convenience API
    • Available Commands
      • Backends
      • Commands
      • Command Identifiers
  • Swift for NNC
nnc
  • Docs »
  • API Reference
  • View page source

API Reference¶

As stated in the Overview, the API designs of NNC is layered. The API reference here breaks down to different levels corresponding to the previous doc.

API reference may contain design comments, which are short-pages extracted from the source code that give high-level overviews of the design motivations.

  • Level 0 API
    • Essentials
  • Level 1 API
    • The concept of meta-ops in Jittor is amazing
    • Tensors
    • Commands
    • Streams
    • Micro Ops
  • Level 2 API
    • Essentials
    • Others
  • Level 3 API
    • Essentials
    • Others
  • Level 3.5 API
    • Construct “switch” control structure in symbolic graph
    • Symbolic graph simplification
    • Construct a “while” loop in a symbolic graph
    • Automatic Differentiation
    • While Loop Essentials
    • Branching
    • Gradient-based Optimization
    • Graph Simplification
    • Automatic Graph Parallelization
    • Memory Compression
    • While Loop Others
  • Level 4 API
    • Essentials
  • Level 5 API
    • What is “dataframe” in ML?
    • Why to support comma-separated-values files in dataframe?
    • Models, layers, and Keras
    • Dataframe API
    • Dataframe Add-ons
    • Dataframe CSV Support
    • Model API
    • Model Add-ons
  • Convenience API
  • Available Commands
    • Backends
    • Commands
    • Command Identifiers
Next Previous

© Copyright 2018, libnnc.org. View the project on GitHub.

Built with Sphinx using a theme provided by Read the Docs.