From ed22e34d654c47017962d3e5758d3a351d8826ab Mon Sep 17 00:00:00 2001
From: 游雁 <zhifu.gzf@alibaba-inc.com>
Date: 星期日, 24 三月 2024 15:03:54 +0800
Subject: [PATCH] finetune

---
 funasr/train_utils/trainer.py |   35 +++++++++++++++++++----------------
 funasr/bin/train.py           |   10 +++++-----
 2 files changed, 24 insertions(+), 21 deletions(-)

diff --git a/funasr/bin/train.py b/funasr/bin/train.py
index e446e54..6cb486b 100644
--- a/funasr/bin/train.py
+++ b/funasr/bin/train.py
@@ -173,11 +173,11 @@
     except:
         writer = None
 
-    if use_ddp or use_fsdp:
-        context = Join([model])
-    else:
-        context = nullcontext()
-
+    # if use_ddp or use_fsdp:
+    #     context = Join([model])
+    # else:
+    #     context = nullcontext()
+    context = nullcontext()
     for epoch in range(trainer.start_epoch, trainer.max_epoch + 1):
         time1 = time.perf_counter()
         with context:
diff --git a/funasr/train_utils/trainer.py b/funasr/train_utils/trainer.py
index e554aca..c665394 100644
--- a/funasr/train_utils/trainer.py
+++ b/funasr/train_utils/trainer.py
@@ -250,14 +250,14 @@
         optim.zero_grad()
         speed_stats = {}
         time5 = time.perf_counter()
-        # iterator_stop = torch.tensor(0).to(self.device)
+        iterator_stop = torch.tensor(0).to(self.device)
 
         dataloader_train.batch_sampler.set_epoch(epoch)
         for batch_idx, batch in enumerate(dataloader_train):
-            # if self.use_ddp or self.use_fsdp:
-            #     dist.all_reduce(iterator_stop, dist.ReduceOp.SUM)
-            #     if iterator_stop > 0:
-            #         break
+            if self.use_ddp or self.use_fsdp:
+                dist.all_reduce(iterator_stop, dist.ReduceOp.SUM)
+                if iterator_stop > 0:
+                    break
             self.batch_total += 1
             time1 = time.perf_counter()
             speed_stats["data_load"] = f"{time1-time5:0.3f}"
@@ -340,7 +340,7 @@
     
                 speed_stats["total_time"] = total_time
                 lr = scheduler.get_last_lr()[0]
-                batch_num_epoch = -1
+                batch_num_epoch = 1
                 if hasattr(dataloader_train, "__len__"):
                     batch_num_epoch = len(dataloader_train)
                 self.log(epoch, batch_idx,
@@ -364,13 +364,15 @@
             if (batch_idx+1) % self.save_checkpoint_interval == 0:
                 self.save_checkpoint(epoch, model=model, optim=optim, scheduler=scheduler, scaler=scaler, step=batch_idx+1)
 
-        # else:
-        #     if self.use_ddp or self.use_fsdp:
-        #         iterator_stop.fill_(1)
-        #         dist.all_reduce(iterator_stop, dist.ReduceOp.SUM)
+        else:
+            if self.use_ddp or self.use_fsdp:
+                iterator_stop.fill_(1)
+                dist.all_reduce(iterator_stop, dist.ReduceOp.SUM)
                 
         if self.use_ddp or self.use_fsdp:
             dist.barrier()
+
+        iterator_stop = torch.tensor(0).to(self.device)
         
         
 
@@ -397,7 +399,7 @@
             
             speed_stats = {}
             time5 = time.perf_counter()
-            # iterator_stop = torch.tensor(0).to(self.device)
+            iterator_stop = torch.tensor(0).to(self.device)
             dataloader_val.batch_sampler.set_epoch(epoch)
             for batch_idx, batch in enumerate(dataloader_val):
                 # if self.use_ddp or self.use_fsdp:
@@ -442,7 +444,7 @@
                     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
                 
-                batch_num_epoch = -1
+                batch_num_epoch = 1
                 if hasattr(dataloader_val, "__len__"):
                     batch_num_epoch = len(dataloader_val)
                 self.log(epoch, batch_idx,
@@ -455,16 +457,17 @@
                          tag="val",
                          )
 
-            # else:
-            #     if self.use_ddp or self.use_fsdp:
-            #         iterator_stop.fill_(1)
-            #         dist.all_reduce(iterator_stop, dist.ReduceOp.SUM)
+            else:
+                if self.use_ddp or self.use_fsdp:
+                    iterator_stop.fill_(1)
+                    dist.all_reduce(iterator_stop, dist.ReduceOp.SUM)
                     
         self.val_acc_list.append(self.val_acc_avg)
         model.train()
 
         if self.use_ddp or self.use_fsdp:
             dist.barrier()
+        iterator_stop = torch.tensor(0).to(self.device)
         
         
     def log(self,

--
Gitblit v1.9.1