tsanley 0.1.0

Tsanley: Understanding Tensor Programs

tsanley

A shape analyzer for tensor programs, using popular tensor libraries: tensorflow, pytorch, numpy, ...

Builds upon a library tsalib for specifying, annotating and transforming tensor shapes using named dimensions.

Quick Start

tsanley discovers shape errors at runtime by checking the runtime tensor shapes with the user-specified shape annotations. Tensor shape annotations are specified in shorthand string format, e.g., x: 'btd'.

More details on shorthand format here.

def func(x):
    x: 'b,t,d' #shape check: ok!
    y: 'b,d' = x.mean(dim=0) #error: dim should be 1
    #   ^ 
    #   | tsanley detects shape violation

def test_func():
    import torch
    from tsalib import get_dim_vars

    B, L, D = get_dim_vars('b t d') # get the declared dimension sizes: 10, 100, 1024
    x = torch.Tensor(B, L, D)
    func(x)

def test():
    #declare the main named dimension variables using tsalib api
    from tsalib import dim_vars
    dim_vars('Batch(b):10 Length(t):100 Hidden(d):1024')

    # initialize tsanley's dynamic shape analyzer
    from tsanley.dynamic import init_analyzer
    init_analyzer(trace_func_names=['func'], show_updates=True)

    test_func()

Installation

pip install tsanley

Status: Experimental

Tested with pytorch examples. tensorflow and numpy programs should also work (tsalib supported backends).