diff --git a/finetune_speaker_v2.py b/finetune_speaker_v2.py
index d466c84..a0709d4 100644
--- a/finetune_speaker_v2.py
+++ b/finetune_speaker_v2.py
@@ -100,10 +100,11 @@ def run(rank, n_gpus, hps):
   # load existing model
   if hps.cont:
       try:
-          _, _, _, epoch_str = utils.load_checkpoint("./pretrained_models/G_latest.pth", net_g, None)
-          _, _, _, epoch_str = utils.load_checkpoint("./pretrained_models/D_latest.pth", net_d, None)
+          _, _, _, epoch_str = utils.load_checkpoint("./OUTPUT_MODEL/G_latest.pth", net_g, None)
+          _, _, _, epoch_str = utils.load_checkpoint("./OUTPUT_MODEL/D_latest.pth", net_d, None)
           global_step = epoch_str * hps.train.batch_size
       except:
+          print("Failed to find latest checkpoint, loading G_0.pth...")
           _, _, _, epoch_str = utils.load_checkpoint("./pretrained_models/G_0.pth", net_g, None)
           _, _, _, epoch_str = utils.load_checkpoint("./pretrained_models/D_0.pth", net_d, None)
           epoch_str = 1