-
I recently removed the pre-processing of mixed-case words (e.g. "HelloWorld") from Orca because from my testing the synthesizers were already doing this work, so there was no need for Orca to do it.
…
-
I am trying to compute reconstruction loss on held-out data, to do model selection.
However, I can't find anywhere in this repo or CTGan repo how to do this. In fact, I can't even find instances of…
-
### Checks
- [X] I have searched the bug with a few keywords, and I confirm this bug was not yet reported.
### Location
Horizon
### Describe the issue
Borg/Drone material synthesizers (glass/stee…
-
Since the release of NVDA 2019.3, there has been a problem that many synthesizers did not work correctly with continuous reading, stopping speech at the end of a sentence.
I have observed that eventu…
-
### Environment Details
Please indicate the following details about the environment in which you found the bug:
* CTGAN version:
* Python version:3.9
* Operating System:windows
### Error De…
-
Case: Running notebook "TimeGAN - Synthetic stock data.ipynb" on Google Collab
I assume because of versioning issues (see below) upon pip installation of ydata-synthetic. But for now it's CPU only.…
-
I have found, you are using MIDI synths are requiring to use external bank files to work, otherwise, the music will be silent. It's a big con of SF2-based and WT synthesizers. I have a suggestion, why…
-
For example:
- delay
- other filter types (high-pass, band-pass, comb filter)
- more efficient saw oscillator
- efficient pulse oscillator (pulse-width modulated)
- lfos (maybe just add a frequency pa…
-
Can we build complex modular synthesizer routings with Python code? Controlling a synthesizer's filter cutoff with an ADSR?
Mockup code:
```python
synth = engine.make_synth_processor()
adsr1…
-
### Bug Description
TokenCountingHandler dies trying to calculate token count (`get_tokens_from_response`) for the response produced by MockLLM.
MockLLM.complete produces `CompletionResponse` with…