Can WebUI print out command for training index as well as inference?

[chigkim]

If you click the button that goes through all the steps, it prints out what command it's executing.

Step 1: Processing data
python3 trainset_preprocess_pipeline_print.py...
step2a:正在提取音高
python3 extract_f0_rmvpe.py...
python3 extract_f0_rmvpe.py...
Step 2b: Extracting features
python3 extract_feature_print.py...
Step 3a: Model training started
python3 train_nsf_sim_cache_sid_load_pretrain.py...
training index
adding index
...
All processes have been completed!

However, it doesn't print out the command for training index.

Same thing, it does not print out the command for inferencing.

Can we have those commands to be printed in both terminal as well as in WebUI?

I haven't checked ckpt processing tab, but it will be great if they print out the commands as well if they don't already.

It's extremely useful to be able to get those commands and run without WebUI.

[elcolex777]

I just separate code for train index in new script (my forked repo). And use it in Kaggle notebook without web-ui. May be some later I will make pull request to original repo.

So you can try to reuse my script: https://github.com/elcolex777/Retrieval-based-Voice-Conversion-WebUI/blob/main/train_index_print.py

I have also upload my kaggle notebook which I use. It’s worked, I think) https://github.com/elcolex777/Retrieval-based-Voice-Conversion-WebUI/blob/main/Retrieval_based_Voice_Conversion_WebUI_v2_kaggle.ipynb

[RVC-Boss]

inference: refer to https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/blob/main/infer_batch_rvc.py or https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/myinfer-v2-0528.py (you may need to modify something)

train the index: refer to https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/blob/main/infer-web.py#L1023 they don't use commands so... maybe you need to learn python or find someone to help you write the command line version

[SECRET-GUEST]

I just made the change for new vc models (works also for olders but not really adapted) I submitted the contribution to hugging face here is the file :

# runtime\python.exe myinfer.py 0 "C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\INPUTS_VOCAL\vocal.wav" "C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\logs\Hagrid.index" harvest "C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\INPUTS_VOCAL\test.wav" "C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\weights\HagridFR.pth" 0.6 cuda:0 True 5 44100 44100 1.0 1.0 True

import os,sys,pdb,torch
now_dir = os.getcwd()
sys.path.append(now_dir)
import argparse
import glob
import sys
import torch
from multiprocessing import cpu_count
class Config:
    def __init__(self,device,is_half):
        self.device = device
        self.is_half = is_half
        self.n_cpu = 0
        self.gpu_name = None
        self.gpu_mem = None
        self.x_pad, self.x_query, self.x_center, self.x_max = self.device_config()

    def device_config(self) -> tuple:
        if torch.cuda.is_available():
            i_device = int(self.device.split(":")[-1])
            self.gpu_name = torch.cuda.get_device_name(i_device)
            if (
                ("16" in self.gpu_name and "V100" not in self.gpu_name.upper())
                or "P40" in self.gpu_name.upper()
                or "1060" in self.gpu_name
                or "1070" in self.gpu_name
                or "1080" in self.gpu_name
            ):
                print("16系/10系显卡和P40强制单精度")
                self.is_half = False
                for config_file in ["32k.json", "40k.json", "48k.json"]:
                    with open(f"configs/{config_file}", "r") as f:
                        strr = f.read().replace("true", "false")
                    with open(f"configs/{config_file}", "w") as f:
                        f.write(strr)
                with open("trainset_preprocess_pipeline_print.py", "r") as f:
                    strr = f.read().replace("3.7", "3.0")
                with open("trainset_preprocess_pipeline_print.py", "w") as f:
                    f.write(strr)
            else:
                self.gpu_name = None
            self.gpu_mem = int(
                torch.cuda.get_device_properties(i_device).total_memory
                / 1024
                / 1024
                / 1024
                + 0.4
            )
            if self.gpu_mem <= 4:
                with open("trainset_preprocess_pipeline_print.py", "r") as f:
                    strr = f.read().replace("3.7", "3.0")
                with open("trainset_preprocess_pipeline_print.py", "w") as f:
                    f.write(strr)
        elif torch.backends.mps.is_available():
            print("没有发现支持的N卡, 使用MPS进行推理")
            self.device = "mps"
        else:
            print("没有发现支持的N卡, 使用CPU进行推理")
            self.device = "cpu"
            self.is_half = True

        if self.n_cpu == 0:
            self.n_cpu = cpu_count()

        if self.is_half:
            # 6G显存配置
            x_pad = 3
            x_query = 10
            x_center = 60
            x_max = 65
        else:
            # 5G显存配置
            x_pad = 1
            x_query = 6
            x_center = 38
            x_max = 41

        if self.gpu_mem != None and self.gpu_mem <= 4:
            x_pad = 1
            x_query = 5
            x_center = 30
            x_max = 32

        return x_pad, x_query, x_center, x_max

f0up_key=sys.argv[1]                     
input_path = sys.argv[2]                 
index_path = sys.argv[3]                 
f0method = sys.argv[4]                   
opt_path = sys.argv[5]                   
model_path = sys.argv[6]                 
index_rate = float(sys.argv[7])          
device = sys.argv[8]                     
is_half = bool(sys.argv[9])              
filter_radius = int(sys.argv[10])        
tgt_sr = int(sys.argv[11])               
resample_sr = int(sys.argv[12])          
rms_mix_rate = float(sys.argv[13])       
version = sys.argv[14]                   
protect = sys.argv[15].lower() == 'false'

print(sys.argv)
config=Config(device,is_half)
now_dir=os.getcwd()
sys.path.append(now_dir)
from vc_infer_pipeline import VC
from lib.infer_pack.models import SynthesizerTrnMs768NSFsid, SynthesizerTrnMs768NSFsid_nono
from lib.audio import load_audio
from fairseq import checkpoint_utils
from scipy.io import wavfile

hubert_model=None
def load_hubert():
    global hubert_model
    models, saved_cfg, task = checkpoint_utils.load_model_ensemble_and_task(["hubert_base.pt"],suffix="",)
    hubert_model = models[0]
    hubert_model = hubert_model.to(device)
    if(is_half):hubert_model = hubert_model.half()
    else:hubert_model = hubert_model.float()
    hubert_model.eval()

def vc_single(sid, input_audio, f0_up_key, f0_file, f0_method, file_index, index_rate, filter_radius, tgt_sr, resample_sr, rms_mix_rate, version, protect):
    global net_g, vc, hubert_model
    if input_audio is None:return "You need to upload an audio", None
    f0_up_key = int(f0_up_key)
    audio=load_audio(input_audio,16000)
    times = [0, 0, 0]
    if(hubert_model==None):load_hubert()
    if_f0 = cpt.get("f0", 1)
    # audio_opt=vc.pipeline(hubert_model,net_g,sid,audio,times,f0_up_key,f0_method,file_index,file_big_npy,index_rate,if_f0,f0_file=f0_file)
    audio_opt = vc.pipeline(hubert_model, net_g, sid, audio, input_path, times, f0_up_key, f0_method, index_path, index_rate, if_f0, filter_radius, tgt_sr, resample_sr, rms_mix_rate, version, protect, f0_file=f0_file)
    print(times)
    return audio_opt

def get_vc(model_path):
    global n_spk,tgt_sr,net_g,vc,cpt,device,is_half
    print("loading pth %s"%model_path)
    cpt = torch.load(model_path, map_location="cpu")
    tgt_sr = cpt["config"][-1]
    cpt["config"][-3]=cpt["weight"]["emb_g.weight"].shape[0]#n_spk
    if_f0=cpt.get("f0",1)
    if(if_f0==1):
        net_g = SynthesizerTrnMs768NSFsid(*cpt["config"], is_half=is_half)
    else:
        net_g = SynthesizerTrnMs768NSFsid_nono(*cpt["config"])
    del net_g.enc_q
    print(net_g.load_state_dict(cpt["weight"], strict=False))  # 不加这一行清不干净，真奇葩
    net_g.eval().to(device)
    if (is_half):net_g = net_g.half()
    else:net_g = net_g.float()
    vc = VC(tgt_sr, config)
    n_spk=cpt["config"][-3]
    # return {"visible": True,"maximum": n_spk, "__type__": "update"}

get_vc(model_path)
wav_opt = vc_single(0, input_path, f0up_key, None, f0method, index_path, index_rate, filter_radius, tgt_sr, resample_sr, rms_mix_rate, version, protect)
wavfile.write(opt_path, tgt_sr, wav_opt)

And here an update of the README.md english to use the file :

## Using the script without a graphical interface (via command line)

For users who prefer a non-GUI approach, you can directly execute the script via the command line.

### Base command:

runtime\python.exe myinfer.py 0 "C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\INPUTS_VOCAL\vocal.wav" "C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\logs\Hagrid.index" harvest "C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\INPUTS_VOCAL\test.wav" "C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\weights\HagridFR.pth" 0.6 cuda:0 True 5 44100 44100 1.0 1.0 True

### Explanation of arguments:

1. **Target voice number**: `0` (in this example)
2. **Path to the input audio file**: `"C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\INPUTS_VOCAL\vocal.wav"`
3. **Path to the index file**: `"C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\logs\Hagrid.index"`
4. **Method for pitch (F0) extraction**: `harvest` (in this example)
5. **Output path for the processed audio file**: `"C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\INPUTS_VOCAL\test.wav"`
6. **Path to the model**: `"C:\ YOUR PATH FOR THE ROOT (RVC0813Nvidia)\weights\HagridFR.pth"`
7. **Index rate**: `0.6` (in this example)
8. **Device for execution (GPU/CPU)**: `cuda:0` for an NVIDIA card, for example.
9. **Copyright protection (True/False)**.
10. **Filter radius**: `5` (in this example)
11. **Target sampling rate**: `44100` (in this example)
12. **Resampling rate**: `44100` (in this example)
13. **RMS mixing rate**: `1.0` (in this example)
14. **Version**: `1.0` (in this example)
15. **Protection**: `True` (in this example)

Ensure you replace paths with the ones relevant to your setup and adjust other parameters as per your requirements.