Fix IndentationError: properly rewrite rvc_logic files and cleanup binaries

rvc_logic/core_train/common.py

@@ -102,4 +102,3 @@ def validate_devices(
             return "cuda", set(validated_devices)
         case DeviceType.CPU:
             return "cpu", None
-None

rvc_logic/core_train/extract.py

@@ -142,5 +142,3 @@ def extract_features(
         f0_method,
         embedder_model_id,
     )
-model_id,
-    )

rvc_logic/core_train/prepare.py

@@ -201,5 +201,3 @@ def preprocess_dataset(
         overlap_len,
         normalization_mode,
     )
-ion_mode,
-    )

rvc_logic/core_train/train.py

@@ -366,4 +366,3 @@ def stop_training(model_name: str) -> None:
         json_dump(updated_training_info_dict, training_info_path)
     except Exception as e:  # noqa: BLE001
         logger.error("Error stopping training: %s", e)  # noqa: TRY400
-s", e)  # noqa: TRY400

rvc_logic/core_train/typing_extra.py

@@ -40,4 +40,3 @@ class TrainingInfo(BaseModel):
     process_pids: list[int] = []
     # TODO add more attributes later
     model_config = ConfigDict(extra="allow")
-ow")

rvc_logic/rvc/configs/config.py

@@ -102,4 +102,3 @@ def get_number_of_gpus():
         num_gpus = torch.cuda.device_count()
         return "-".join(map(str, range(num_gpus)))
     return "-"
-"

rvc_logic/rvc/infer/infer.py

@@ -525,4 +525,3 @@ class VoiceConverter:
         if self.cpt is not None:
             self.vc = VC(self.tgt_sr, self.config)
             self.n_spk = self.cpt["config"][-3]
-f.cpt["config"][-3]

rvc_logic/rvc/infer/pipeline.py

@@ -578,4 +578,3 @@ class Pipeline:
         if torch.cuda.is_available():
             torch.cuda.empty_cache()
         return audio_opt
-t

rvc_logic/rvc/infer/typing_extra.py

@@ -55,4 +55,3 @@ class ConvertAudioKwArgs(TypedDict, total=False):
     delay_seconds: float
     delay_feedback: float
     delay_mix: float
-t

rvc_logic/rvc/lib/algorithm/attentions.py

@@ -255,4 +255,3 @@ class FFN(torch.nn.Module):
             x,
             convert_pad_shape([[0, 0], [0, 0], [pad, pad]]),
         )
-)

rvc_logic/rvc/lib/algorithm/discriminators.py

@@ -264,4 +264,3 @@ class DiscriminatorR(torch.nn.Module):
         mag = torch.norm(torch.view_as_real(x), p=2, dim=-1)  # [B, F, TT]
 
         return mag
- mag

rvc_logic/rvc/lib/algorithm/encoders.py

@@ -225,4 +225,3 @@ class PosteriorEncoder(torch.nn.Module):
             ):
                 torch.nn.utils.remove_weight_norm(self.enc)
         return self
-eturn self

rvc_logic/rvc/lib/algorithm/generators/hifigan.py

@@ -246,4 +246,3 @@ class SineGenerator(torch.nn.Module):
             sine_waveforms = sine_waves * voiced_mask + noise
 
         return sine_waveforms, voiced_mask, noise
-oise

rvc_logic/rvc/lib/algorithm/generators/hifigan_nsf.py

@@ -255,4 +255,3 @@ class HiFiGANNSFGenerator(torch.nn.Module):
                 ):
                     remove_weight_norm(l)
         return self
-rn self

rvc_logic/rvc/lib/algorithm/generators/refinegan.py

@@ -459,4 +459,3 @@ class RefineGANGenerator(nn.Module):
 
         for block in self.upsample_conv_blocks:
             block.remove_weight_norm()
-)

rvc_logic/rvc/lib/algorithm/modules.py

@@ -117,4 +117,3 @@ class WaveNet(torch.nn.Module):
             torch.nn.utils.remove_weight_norm(layer)
         for layer in self.res_skip_layers:
             torch.nn.utils.remove_weight_norm(layer)
-)

rvc_logic/rvc/lib/algorithm/residuals.py

@@ -268,4 +268,3 @@ class ResidualCouplingLayer(torch.nn.Module):
 
     def remove_weight_norm(self):
         self.enc.remove_weight_norm()
-rm()

rvc_logic/rvc/lib/algorithm/synthesizers.py

@@ -248,4 +248,3 @@ class Synthesizer(torch.nn.Module):
         )
 
         return o, x_mask, (z, z_p, m_p, logs_p)
-, z_p, m_p, logs_p)

rvc_logic/rvc/lib/predictors/F0Extractor.py

@@ -108,4 +108,3 @@ class F0Extractor:
         F_temp[F_temp == 0] = np.nan
         F_cents = 1200 * np.log2(F_temp / F_ref)
         return F_cents
-F_cents

rvc_logic/rvc/lib/predictors/f0.py

@@ -89,4 +89,3 @@ class FCPE:
         )
 
         return f0
-n f0

rvc_logic/rvc/lib/tools/model_download.py

@@ -235,4 +235,3 @@ def clean_extracted_files(extract_folder_path, model_name):
         destination_path = os.path.join(extract_folder_path, new_file_name)
         if not pathlib.Path(destination_path).exists():
             pathlib.Path(source_path).rename(destination_path)
-ath)

rvc_logic/rvc/lib/tools/prerequisites_download.py

@@ -195,4 +195,3 @@ def prequisites_download_pipeline(
             if pretraineds_hifigan:
                 download_mapping_files(pretraineds_hifigan_list, global_bar)
                 download_mapping_files(pretraineds_refinegan_list, global_bar)
-)

rvc_logic/rvc/lib/tools/pretrained_selector.py

@@ -13,4 +13,3 @@ def pretrained_selector(vocoder: str, sample_rate: int) -> tuple[str, str]:
     if pathlib.Path(path_g).exists() and pathlib.Path(path_d).exists():
         return path_g, path_d
     return "", ""
-"

rvc_logic/rvc/lib/utils.py

@@ -191,4 +191,3 @@ def load_embedding(embedder_model, custom_embedder=None):
 
     models = HubertModelWithFinalProj.from_pretrained(model_path)
     return models
-s

rvc_logic/rvc/train/data_utils.py

@@ -393,4 +393,3 @@ class DistributedBucketSampler(torch.utils.data.distributed.DistributedSampler):
         Returns the length of the sampler.
         """
         return self.num_samples // self.batch_size
-size

rvc_logic/rvc/train/extract/extract.py

@@ -258,4 +258,3 @@ def update_model_info(
     data["custom_embedder_model_hash"] = custom_embedder_model_hash
     with pathlib.Path(file_path).open("w") as f:
         json.dump(data, f, indent=4)
-ta, f, indent=4)

rvc_logic/rvc/train/extract/preparing_files.py

@@ -112,4 +112,3 @@ def generate_filelist(
 
     with pathlib.Path(os.path.join(model_path, "filelist.txt")).open("w") as f:
         f.write("\n".join(options))
-ns))

rvc_logic/rvc/train/preprocess/preprocess.py

@@ -432,5 +432,3 @@ def preprocess_training_set(
         elapsed_time,
         format_duration(audio_length),
     )
-h),
-    )

rvc_logic/rvc/train/train.py

@@ -1110,4 +1110,3 @@ def cleanup_training_processes(experiment_dir) -> None:
     with pathlib.Path(pid_file_path).open("w") as pid_file:
         pid_data.pop("process_pids", None)
         json.dump(pid_data, pid_file, indent=4)
-dump(pid_data, pid_file, indent=4)