From 46ec82c05410b600b0af33b71c4f1c818a46c698 Mon Sep 17 00:00:00 2001
From: 游雁 <zhifu.gzf@alibaba-inc.com>
Date: 星期二, 26 三月 2024 16:47:10 +0800
Subject: [PATCH] update
---
funasr/train_utils/trainer.py | 6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/funasr/train_utils/trainer.py b/funasr/train_utils/trainer.py
index 2d47fc1..56ec604 100644
--- a/funasr/train_utils/trainer.py
+++ b/funasr/train_utils/trainer.py
@@ -190,7 +190,7 @@
if self.resume:
ckpt = os.path.join(self.output_dir, "model.pt")
if os.path.isfile(ckpt):
- checkpoint = torch.load(ckpt)
+ checkpoint = torch.load(ckpt, map_location="cpu")
self.start_epoch = checkpoint['epoch'] + 1
# self.model.load_state_dict(checkpoint['state_dict'])
src_state = checkpoint['state_dict']
@@ -215,7 +215,7 @@
self.val_acc_list = checkpoint["acc"]
self.step_or_epoch = checkpoint["step_or_epoch"]
-
+ model.to(self.device)
print(f"Checkpoint loaded successfully from '{ckpt}'")
else:
print(f"No checkpoint found at '{ckpt}', does not resume status!")
@@ -440,7 +440,7 @@
dist.all_reduce(val_acc_avg, op=dist.ReduceOp.SUM)
self.val_loss_avg = val_loss_avg.detach().cpu().item() / self.world_size
self.val_acc_avg = val_acc_avg.detach().cpu().item() / self.world_size
-
+ time5 = time.perf_counter()
batch_num_epoch = 1
if hasattr(dataloader_val, "__len__"):
batch_num_epoch = len(dataloader_val)
--
Gitblit v1.9.1