Update README.md

This commit is contained in:
Anjok07 2021-05-06 03:56:19 -05:00 committed by GitHub
parent a6e4d99eea
commit bddef0d785
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -103,6 +103,24 @@ python 12_model_ens_inference.py -g 0 -s -i "INPUT"
- **Please Note:** Do not specify the model parameters or a model for the ensemble inference scripts! Those details are already fixed within the script. All final outputs generated through the ensemble scripts can be found in the "ensembled" folder.
### Ensembler
The ensembler has the ability to take 2 instrummental or vocal outputs generated by different models and combine the best results from both! Here is how to use it manually -
- For instrumental outputs, run the following command:
```
python lib/spec_utils.py -a min_mag -m modelparams/1band_sr44100_hl512.json "INPUT1" "INPUT2" -o "CHOOSEFILENAME"
```
- For instrumental outputs, run the following command:
```
python lib/spec_utils.py -a max_mag -m modelparams/1band_sr44100_hl512.json "INPUT1" "INPUT2" -o "CHOOSEFILENAME"
```
To automate the ensembling process, please use the 'Ensembler-2to9-Outputs.bat' script reference below.
### Windows Batch Files
We included the following Windows batch files to help automate commands:
@ -117,6 +135,7 @@ We included the following Windows batch files to help automate commands:
- Once you hit enter, you will be asked to enter the number of files dropped into the batch script.
- The ensembled outputs can be found in the "ensembled" folder.
- **This script is going to be updated soon so that you won't have to manually enter the number of files dropped**
- **The ensembler will also be updated as a python script for cross platform compatibility as well.**
## Troubleshooting