update

examples/dtln/step_2_train_model.py

@@ -1,7 +1,8 @@
 #!/usr/bin/python3
 # -*- coding: utf-8 -*-
 """
-https://github.com/Rikorose/DeepFilterNet
+https://github.com/breizhn/DTLN
+
 """
 import argparse
 import json

examples/rnnoise/step_2_train_model.py

@@ -194,18 +194,12 @@ def main():
     if last_step_idx != -1:
         logger.info(f"resume from steps-{last_step_idx}.")
         model_pt = serialization_dir / f"steps-{last_step_idx}/model.pt"
-        optimizer_pth = serialization_dir / f"steps-{last_step_idx}/optimizer.pth"
 
         logger.info(f"load state dict for model.")
         with open(model_pt.as_posix(), "rb") as f:
             state_dict = torch.load(f, map_location="cpu", weights_only=True)
         model.load_state_dict(state_dict, strict=True)
 
-        logger.info(f"load state dict for optimizer.")
-        with open(optimizer_pth.as_posix(), "rb") as f:
-            state_dict = torch.load(f, map_location="cpu", weights_only=True)
-        optimizer.load_state_dict(state_dict)
-
     if config.lr_scheduler == "CosineAnnealingLR":
         lr_scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(
             optimizer,
@@ -265,7 +259,7 @@ def main():
 
         progress_bar_train = tqdm(
             initial=step_idx,
-            desc="Training; epoch: {}".format(epoch_idx),
+            desc="Training; epoch-{}".format(epoch_idx),
         )
         for train_batch in train_data_loader:
             clean_audios, noisy_audios = train_batch