various enhancements

[koke2c95]

great works!

debug these math expression code is a very big problem

not only scalar vector matrix , high rank tensor support

Visualizing 3D tensors and beyond:

TensorDim is contain : left to right [N,C,H,W] N->C->H->W
TensorDim like flatten Onions,cabbage box in box represent dim
TensorDim has real represention information (N,C,H,W etc.)

namedtensor support

plaidml dsl
Types Tensor TensorDim TensorIndex

represent actual data:
audio, 1D plot, text : vector
image ,2D plot : matrix
etc.

expression graph:

different color input variable ,leaf parameters/variable
with temporal variable / activation : "Road width"
elementwise leaf / elementwise or slice or other connection(edge) print
AST forward & backward mode

animation:

expression step computation mode (debug)
slice ,reshape, .T() other manipulate N-d array TensorDim/TensorIndex operators
example : matmul = m@v
m@v = (v.squeeze(0).expand_as(w) * w).sum(1,keepdim=True).unsqueeze(1)

more ... conv2d etc

interactive:

interactive build block(visual programing)
reverse interactive(debug) : selected tensor elements and follow it in expression graph
selected element : muti view , like convolution-visualizer Input(Inputgrad) Weight(Weightgrad) Output(Outputgrad) views

NN support:

NN module visualization (conv2d)
bigger computation graph : pytorchrec
multi computation graph visualization and live debug

other useful link

SimpNet memory map

Visualize the virtual address space of a Windows process on a Hilbert curve.

nvidia ppt

tensor network

einops

[parrt]

hi! thanks for all the ideas and links. have you ever seen a good basic library for animating things like matrices?

[koke2c95]

hi! thanks for all the ideas and links. have you ever seen a good basic library for animating things like matrices?

Sorry for the late reply to you.
I search a lot , but I haven't seen visualization library like this idea.

now

back to basics.

The Ndarray/Tensor is the thing that stores data, values
visualization by
Think of each value as a pixel
And those tensor operations are essentially is pixel coordinates in the operating space
it can shuffle, drop, add any pixel

previous

you can make a visualizing by nest different type (dim Information & tensor data binding)
example like. muti same image nest in a matrix eq. grid image
and you can infinity nesting

(0,0) = R scalar
(0,1) = R vector
(0,2) = R matrix
(0,~) ~>=3 : R tensor
(1,2) = gray image
(3,2) = rgb image
(0,1,2) = 2 dim R field R^2
(0,2,2) = 2 dim C field C^2