Göran Sandström

RN it's very unclear how we're actually supposed to get a model to use with ctcdecode I compile KenLM separately and do kenlm/bin/lmplz etc, but it seems to me that...

Add ability to supply extra scorer

1

It would be great to have the ability to supply an extra scorer (or several) such as a transformer decoder, CNN-LM, etc. This way, one can combine multiple methods to...

Is it using fast-wavenet-decoding?

11

I'm trying to figure out if fast-wavenet-decoding (https://github.com/tomlepaine/fast-wavenet) is used when decoding, if not, it's quite simple to implement and would yield a big performance boost, of couse IAF-decoder would...

So, when playing with vectalign I noticed that it fails a lot of the time when trying to morph between assets where the "centerline", that is, the visual rotation axis...

Griffin lim reconstruction code source

1

NSynth uses griffin lim reconstruction, with a nice implementation in their utils class https://github.com/tensorflow/magenta/blob/master/magenta/models/nsynth/utils.py

0xThings - Distributed manufacturing district

3

## **Name**: # 0xThings ## **Purpose**: Purchase 3D things modelled by other users printed locally and sent to you via mail. ## **Description**: There are many owners of 3D printers...

Some questions regarding compound: Is there any limit on what kind of capabilities I can declare? How do I declare a ternary capabilities converter correctly?

Home assistant is really great at monitoring media players, it would be nice to be able to add it as a speaker device (playing, next, previous, etc)

When extracting simple info like width and height from an MP4File hachoir seems to first parse out the entire file before yielding the info. Dimensions and probably duration too, should...

TL;DR make it possible to use tortoise-orm as a vector database **Is your feature request related to a problem? Please describe.** A lot of commercial vector databases are being thrown...