theflyingzamboni

Results 33 issues of theflyingzamboni

When using the Resample option in the Scale tab (but not Scale option) with any of the (Pixel) Target Scale options, the scale input is still treated as a percentage...

## Environment data - Language Server version: 2022.7.40 - OS and version: Windows 10 Home - Python version: 3.9.7 ## Code Snippet (using onnx=1.11) ```python import onnx model = onnx.ModelProto()...

needs investigation
needs stub

In The Legend of Dragoon (and likely other games, though haven't tested personally), textures with multiple CLUTs dump as multiple images, one for each CLUT. This results in multiple (sometimes...

## Description The issue is twofold: 1) Certain textures in the game do not dump properly (or at all). This primarily seems to occur with animated textures in otherwise 2D...

Shift node (and possibly others with negative-space fills as well) output typing does not account for transparency fill expanding channels to 4. For instance, inputting a 1-channel image to Shift...

bug

Allow individual nodes to be disabled and permit upstream inputs to be passed through if possible, else disable downstream nodes as well. One way this could potentially work at a...

enhancement

Running the Preview Image node would pop up a custom chakra window. This custom window would have the following features, as shown in the mockup below: 1) Window should be...

enhancement

Many types of models make use of the Shape and Gather ops, but these still are not supported by ncnn. Using things like onnxoptimizer/onnxsim cannot always optimize these ops a...

The operators.md and operation-param-weight.md files are missing documentation on some ncnn operations. There may be more, but this at least includes: - EmbedLayerNormalization - SkipLayerNormalization

**Motivation** For nodes like Resize where the user can adjust multiple dimensions, it is more convenient to be able to link inputs together to keep them the same/having the same...

enhancement