vanja/neural-amp-modeler-lv2
1
0
Fork 0
mirror of https://github.com/mikeoliphant/neural-amp-modeler-lv2.git synced 2026-05-06 19:50:11 +02:00
Code Issues Packages Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
Mike Oliphant
2023-06-08 11:03:43 -07:00
committed by GitHub
parent 47f75ccca0
commit f1077aca82
1 changed files with 11 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
models/README.md
+11
View File
@@ -1 +1,12 @@
These are some sample NAM models designed to be used for performance testing on CPU-limited devices. They are all based on a [LiveSpice model of a Boss SD-1 pedal](https://blog.nostatic.org/2023/04/this-boss-sd-1-pedal-does-not-exist.html). These are some sample NAM models designed to be used for performance testing on CPU-limited devices. They are all based on a [LiveSpice model of a Boss SD-1 pedal](https://blog.nostatic.org/2023/04/this-boss-sd-1-pedal-does-not-exist.html).
All models were trained for 300 epochs using the NAM "v1_1_1.wav" capture signal.
| Model | ESR | CPU%<br>(RPi4 64bit) | Notes |
| --- |--- | :-: | --- |
| [BossWN-feather.nam](https://github.com/mikeoliphant/neural-amp-modeler-lv2/blob/main/models/BossWN-feather.nam) | .0001 | 37% | WaveNet "feather" preset |
| [BossWN-4x2x1.nam](https://github.com/mikeoliphant/neural-amp-modeler-lv2/blob/main/models/BossWN-4x2x1.nam) | .0003 | 28% | WaveNet 4x2 channel |
| [BossLSTM-2x16.nam](https://github.com/mikeoliphant/neural-amp-modeler-lv2/blob/main/models/BossLSTM-2x16.nam) | .0013 | 28% | LSTM 2x16 layers |
| [BossLSTM-1x24.nam](https://github.com/mikeoliphant/neural-amp-modeler-lv2/blob/main/models/BossLSTM-1x24.nam) | .0017 | 22% | LSTM 1x24 layer |
| [BossLSTM-2x8.nam](https://github.com/mikeoliphant/neural-amp-modeler-lv2/blob/main/models/BossLSTM-2x8.nam) | .0019 | 17% | LSTM 2x8 layers |
| [BossLSTM-1x16.nam](https://github.com/mikeoliphant/neural-amp-modeler-lv2/blob/main/models/BossLSTM-1x16.nam) | .0041 | 15% | LSTM 1x16 layer |