fix: handle exceptions directly in UI scripts Pt.3

Author: Vidalnt

core.py

@@ -438,7 +438,10 @@ def run_preprocess_script(
             ],
         ),
     ]
-    subprocess.run(command)
+    result = subprocess.run(command)
+    if result.returncode != 0:
+        return f"Preprocessing failed for model {model_name}. Please check the console logs for more details."
+
     return f"Model {model_name} preprocessed successfully."
 
 
@@ -474,7 +477,9 @@ def run_extract_script(
         ),
     ]
 
-    subprocess.run(command_1)
+    result = subprocess.run(command_1)
+    if result.returncode != 0:
+        return f"Feature extraction failed for model {model_name}. Please check the console logs for more details."
 
     return f"Model {model_name} extracted successfully."
 
@@ -543,9 +548,8 @@ def run_train_script(
     ]
     result = subprocess.run(command)
     if result.returncode != 0:
-        raise RuntimeError(
-            f"Training failed for model {model_name}. Please check the console logs for more details."
-        )
+        return f"Training failed for model {model_name}. Please check the console logs for more details."
+
     run_index_script(model_name, index_algorithm)
     return f"Model {model_name} trained successfully."
 
@@ -560,7 +564,10 @@ def run_index_script(model_name: str, index_algorithm: str):
         index_algorithm,
     ]
 
-    subprocess.run(command)
+    result = subprocess.run(command)
+    if result.returncode != 0:
+        return f"Index generation failed for model {model_name}. Make sure you have enough GPU available to generate the Index file. Please check the console logs for more details."
+
     return f"Index file for {model_name} generated successfully."
 
 
@@ -587,7 +594,7 @@ def run_tensorboard_script():
 def run_download_script(model_link: str):
     result = model_download_pipeline(model_link)
     if result == "Error" or result is None:
-        return "An error occurred downloading the model. Check the console for details."
+        return "An error occurred downloading the model. Please check the console logs for more details."
     return "Model downloaded successfully."
 
 

rvc/train/extract/extract.py

@@ -1,25 +1,24 @@
+import concurrent.futures
+import glob
+import json
+import multiprocessing as mp
 import os
 import sys
-import glob
 import time
-import tqdm
-import torch
-import torchcrepe
+
 import numpy as np
-import concurrent.futures
-import multiprocessing as mp
-import json
+import torch
+import tqdm
 
 now_dir = os.getcwd()
 sys.path.append(os.path.join(now_dir))
 
 # Zluda hijack
 import rvc.lib.zluda
-
+from rvc.configs.config import Config
+from rvc.lib.predictors.f0 import CREPE, FCPE, RMVPE
 from rvc.lib.utils import load_audio_16k, load_embedding
 from rvc.train.extract.preparing_files import generate_config, generate_filelist
-from rvc.lib.predictors.f0 import CREPE, FCPE, RMVPE
-from rvc.configs.config import Config
 
 # Load config
 config = Config()
@@ -184,6 +183,13 @@ def run_embedding_extraction(
     include_mutes = int(sys.argv[8]) if len(sys.argv) > 8 else 2
 
     wav_path = os.path.join(exp_dir, "sliced_audios_16k")
+
+    if not os.path.exists(wav_path):
+        print(
+            f"Folder for feature extraction not found at {wav_path}. Did you run the preprocessing step?"
+        )
+        sys.exit(1)
+
     os.makedirs(os.path.join(exp_dir, "f0"), exist_ok=True)
     os.makedirs(os.path.join(exp_dir, "f0_voiced"), exist_ok=True)
     os.makedirs(os.path.join(exp_dir, "extracted"), exist_ok=True)
@@ -212,6 +218,12 @@ def run_embedding_extraction(
         ]
         files.append(file_info)
 
+    if not files:
+        print(
+            f"Sliced audios not found at {wav_path}. Did you run the preprocessing step?"
+        )
+        sys.exit(1)
+
     devices = ["cpu"] if gpus == "-" else [f"cuda:{idx}" for idx in gpus.split("-")]
 
     run_pitch_extraction(files, devices, f0_method, num_processes)

rvc/train/preprocess/preprocess.py

@@ -1,26 +1,27 @@
+import concurrent.futures
+import json
+import multiprocessing
 import os
 import sys
 import time
-from scipy import signal
-from scipy.io import wavfile
-import numpy as np
-import concurrent.futures
-from tqdm import tqdm
-import json
 from distutils.util import strtobool
+
 import librosa
-import multiprocessing
 import noisereduce as nr
+import numpy as np
 import soxr
+from scipy import signal
+from scipy.io import wavfile
+from tqdm import tqdm
 
 now_directory = os.getcwd()
 sys.path.append(now_directory)
 
+import logging
+
 from rvc.lib.utils import load_audio
 from rvc.train.preprocess.slicer import Slicer
 
-import logging
-
 logging.getLogger("numba.core.byteflow").setLevel(logging.WARNING)
 logging.getLogger("numba.core.ssa").setLevel(logging.WARNING)
 logging.getLogger("numba.core.interpreter").setLevel(logging.WARNING)
@@ -281,6 +282,13 @@ def preprocess_training_set(
     overlap_len: float,
     normalization_mode: str,
 ):
+    if not os.path.exists(input_root):
+        print(f"The dataset path does not exist: '{input_root}'.")
+        sys.exit(1)
+
+    if not os.path.isdir(input_root):
+        print(f"The dataset path is not a directory: '{input_root}'.")
+        sys.exit(1)
     start_time = time.time()
     pp = PreProcess(sr, exp_dir)
     print(f"Starting preprocess with {num_processes} processes...")
@@ -301,6 +309,11 @@ def preprocess_training_set(
             )
 
     # print(f"Number of files: {len(files)}")
+    if len(files) == 0:
+        print(
+            f"No audio files found in the dataset path: '{input_root}'. Please check that the path is correct and contains valid audio files."
+        )
+        sys.exit(1)
     audio_length = []
     with tqdm(total=len(files)) as pbar:
         with concurrent.futures.ProcessPoolExecutor(

rvc/train/process/extract_index.py

@@ -10,68 +10,68 @@
 exp_dir = str(sys.argv[1])
 index_algorithm = str(sys.argv[2])
 
-try:
-    feature_dir = os.path.join(exp_dir, f"extracted")
-    model_name = os.path.basename(exp_dir)
+feature_dir = os.path.join(exp_dir, "extracted")
+model_name = os.path.basename(exp_dir)
 
-    if not os.path.exists(feature_dir):
-        print(
-            f"Feature to generate index file not found at {feature_dir}. Did you run preprocessing and feature extraction steps?"
-        )
-        sys.exit(1)
+if not os.path.exists(feature_dir):
+    print(
+        f"Feature to generate index file not found at {feature_dir}. Did you run preprocessing and feature extraction steps?"
+    )
+    sys.exit(1)
 
-    index_filename_added = f"{model_name}.index"
-    index_filepath_added = os.path.join(exp_dir, index_filename_added)
+index_filename_added = f"{model_name}.index"
+index_filepath_added = os.path.join(exp_dir, index_filename_added)
 
-    if os.path.exists(index_filepath_added):
-        pass
-    else:
-        npys = []
-        listdir_res = sorted(os.listdir(feature_dir))
+if os.path.exists(index_filepath_added):
+    pass
+else:
+    npys = []
+    print(f"Generating index for '{model_name}', this may take a while...")
+    listdir_res = sorted(os.listdir(feature_dir))
 
-        for name in listdir_res:
-            file_path = os.path.join(feature_dir, name)
-            phone = np.load(file_path)
-            npys.append(phone)
+    for name in listdir_res:
+        file_path = os.path.join(feature_dir, name)
+        phone = np.load(file_path)
+        npys.append(phone)
 
-        big_npy = np.concatenate(npys, axis=0)
+    if not npys:
+        print(
+            f"Feature files in {feature_dir} could not be loaded correctly. Did you run preprocessing and feature extraction steps?"
+        )
+        sys.exit(1)
 
-        big_npy_idx = np.arange(big_npy.shape[0])
-        np.random.shuffle(big_npy_idx)
-        big_npy = big_npy[big_npy_idx]
+    big_npy = np.concatenate(npys, axis=0)
 
-        if big_npy.shape[0] > 2e5 and (
-            index_algorithm == "Auto" or index_algorithm == "KMeans"
-        ):
-            big_npy = (
-                MiniBatchKMeans(
-                    n_clusters=10000,
-                    verbose=True,
-                    batch_size=256 * cpu_count(),
-                    compute_labels=False,
-                    init="random",
-                )
-                .fit(big_npy)
-                .cluster_centers_
-            )
+    big_npy_idx = np.arange(big_npy.shape[0])
+    np.random.shuffle(big_npy_idx)
+    big_npy = big_npy[big_npy_idx]
 
-        n_ivf = min(int(16 * np.sqrt(big_npy.shape[0])), big_npy.shape[0] // 39)
+    if big_npy.shape[0] > 2e5 and (
+        index_algorithm == "Auto" or index_algorithm == "KMeans"
+    ):
+        big_npy = (
+            MiniBatchKMeans(
+                n_clusters=10000,
+                verbose=True,
+                batch_size=256 * cpu_count(),
+                compute_labels=False,
+                init="random",
+            )
+            .fit(big_npy)
+            .cluster_centers_
+        )
 
-        # index_added
-        index_added = faiss.index_factory(768, f"IVF{n_ivf},Flat")
-        index_ivf_added = faiss.extract_index_ivf(index_added)
-        index_ivf_added.nprobe = 1
-        index_added.train(big_npy)
+    n_ivf = min(int(16 * np.sqrt(big_npy.shape[0])), big_npy.shape[0] // 39)
 
-        batch_size_add = 8192
-        for i in range(0, big_npy.shape[0], batch_size_add):
-            index_added.add(big_npy[i : i + batch_size_add])
+    # index_added
+    index_added = faiss.index_factory(768, f"IVF{n_ivf},Flat")
+    index_ivf_added = faiss.extract_index_ivf(index_added)
+    index_ivf_added.nprobe = 1
+    index_added.train(big_npy)
 
-        faiss.write_index(index_added, index_filepath_added)
-        print(f"Saved index file '{index_filepath_added}'")
+    batch_size_add = 8192
+    for i in range(0, big_npy.shape[0], batch_size_add):
+        index_added.add(big_npy[i : i + batch_size_add])
 
-except Exception as error:
-    print(f"An error occurred extracting the index: {error}")
-    print(
-        "If you are running this code in a virtual environment, make sure you have enough GPU available to generate the Index file."
-    )
+    faiss.write_index(index_added, index_filepath_added)
+    print(f"Saved index file '{index_filepath_added}'")

tabs/inference/inference.py

@@ -1173,7 +1173,7 @@ def enforce_terms(terms_accepted, *args):
             except Exception:
                 traceback.print_exc()
                 return (
-                    "An error occurred during audio conversion. Check the console for details.",
+                    "An error occurred during audio conversion. Please check the console logs for more details.",
                     None,
                 )
 
@@ -1186,7 +1186,7 @@ def enforce_terms_batch(terms_accepted, *args):
                 return run_batch_infer_script(*args)
             except Exception:
                 traceback.print_exc()
-                return "An error occurred during audio batch conversion. Check the console for details."
+                return "An error occurred during audio batch conversion. Please check the console logs for more details."
 
         terms_checkbox = gr.Checkbox(
             label=i18n("I agree to the terms of use"),

tabs/train/train.py

@@ -768,10 +768,7 @@ def enforce_terms(terms_accepted, *args):
                 message = "You must agree to the Terms of Use to proceed."
                 gr.Info(message)
                 return message
-            try:
-                return run_train_script(*args)
-            except Exception as e:
-                return e
+            return run_train_script(*args)
 
         terms_checkbox = gr.Checkbox(
             label=i18n("I agree to the terms of use"),

tabs/tts/tts.py

@@ -363,7 +363,7 @@ def enforce_terms(terms_accepted, *args):
         except Exception:
             traceback.print_exc()
             return (
-                "An error occurred during TTS conversion. Check the console for details.",
+                "An error occurred during TTS conversion. Please check the console logs for more details.",
                 None,
             )