From 2ac38adbe5f4e1374a079e032ed4b504351a207c Mon Sep 17 00:00:00 2001
From: zhifu gao <zhifu.gzf@alibaba-inc.com>
Date: 星期二, 23 四月 2024 18:08:57 +0800
Subject: [PATCH] Dev gzf exp (#1647)
---
funasr/train_utils/trainer.py | 163 +++++++++++++++++++++++++++++++++---------------------
1 files changed, 100 insertions(+), 63 deletions(-)
diff --git a/funasr/train_utils/trainer.py b/funasr/train_utils/trainer.py
index d0023fd..3ee6885 100644
--- a/funasr/train_utils/trainer.py
+++ b/funasr/train_utils/trainer.py
@@ -71,21 +71,19 @@
self.use_ddp = use_ddp
self.use_fsdp = use_fsdp
self.device = kwargs.get('device', "cuda")
- self.avg_nbest_model = kwargs.get("avg_nbest_model", 5)
# self.kwargs = kwargs
self.log_interval = kwargs.get("log_interval", 50)
self.batch_total = 0
self.use_fp16 = use_fp16
- self.disable_gpu_cache = kwargs.get("disable_gpu_cache", True)
- # scaler = GradScaler(enabled=use_fp16) if use_fp16 else None
- # scaler = ShardedGradScaler(enabled=use_fp16) if use_fsdp else scaler
- # self.scaler = scaler
self.save_checkpoint_interval = kwargs.get("save_checkpoint_interval", 5000)
- self.keep_nbest_models = kwargs.get("keep_nbest_models", -1)
+ self.validate_interval = kwargs.get("validate_interval", 5000)
+ self.keep_nbest_models = kwargs.get("keep_nbest_models", 500)
+ self.avg_keep_nbest_models_type = kwargs.get("avg_keep_nbest_models_type", "acc")
+ self.avg_nbest_model = kwargs.get("avg_nbest_model", 10)
self.accum_grad = kwargs.get("accum_grad", 1)
self.grad_clip = kwargs.get("grad_clip", 10.0)
self.grad_clip_type = kwargs.get("grad_clip_type", 2.0)
- self.validate_interval = kwargs.get("validate_interval", 5000)
+
try:
@@ -103,8 +101,10 @@
self.val_loss_avg = 0.0
self.best_acc_idx = 0
self.saved_ckpts = {}
- self.val_acc_list = []
self.step_or_epoch = -1
+ self.best_step_or_epoch = ""
+ self.val_acc_step_or_eoch = {}
+ self.val_loss_step_or_eoch = {}
def save_checkpoint(self, epoch,
step=None,
@@ -124,14 +124,17 @@
if self.rank == 0:
logging.info(f"Save checkpoint: {epoch}, rank: {self.local_rank}\n")
- self.step_or_epoch += 1
+ # self.step_or_epoch += 1
state = {
'epoch': epoch,
'state_dict': model.state_dict(),
'optimizer': optim.state_dict(),
'scheduler': scheduler.state_dict(),
- "acc": self.val_acc_list,
- "step_or_epoch": self.step_or_epoch,
+ "saved_ckpts": self.saved_ckpts,
+ "val_acc_step_or_eoch": self.val_acc_step_or_eoch,
+ "val_loss_step_or_eoch": self.val_loss_step_or_eoch,
+ "best_step_or_epoch": self.best_step_or_epoch,
+ "avg_keep_nbest_models_type": self.avg_keep_nbest_models_type,
}
if hasattr(model, "module"):
state["state_dict"] = model.module.state_dict()
@@ -150,23 +153,37 @@
logging.info(f'\nCheckpoint saved to {filename}\n')
latest = Path(os.path.join(self.output_dir, f'model.pt'))
torch.save(state, latest)
-
- if self.val_acc_list[self.step_or_epoch] >= self.val_acc_list[self.best_acc_idx]:
- self.best_acc_idx = self.step_or_epoch
- best_ckpt = Path(os.path.join(self.output_dir, f'model.pt.best'))
- torch.save(state, best_ckpt)
- logging.info(f"Update best acc: {self.val_acc_list[self.best_acc_idx]}, {best_ckpt}")
+ if self.best_step_or_epoch == "":
+ self.best_step_or_epoch = ckpt_name
+
+ if self.avg_keep_nbest_models_type == "acc":
+ if self.val_acc_step_or_eoch[ckpt_name] >= self.val_acc_step_or_eoch[self.best_step_or_epoch]:
+ self.best_step_or_epoch = ckpt_name
+ best_ckpt = Path(os.path.join(self.output_dir, f'model.pt.best'))
+ torch.save(state, best_ckpt)
+ logging.info(f"Update best acc: {self.val_acc_step_or_eoch[self.best_step_or_epoch]:.4f}, {best_ckpt}")
+ else:
+ logging.info(f"No improvement in acc: {self.val_acc_step_or_eoch[ckpt_name]:.4f} < {self.val_acc_step_or_eoch[self.best_step_or_epoch]:.4f}")
+ elif self.avg_keep_nbest_models_type == "loss":
+ if self.val_loss_step_or_eoch[ckpt_name] <= self.val_loss_step_or_eoch[self.best_step_or_epoch]:
+ self.best_step_or_epoch = ckpt_name
+ best_ckpt = Path(os.path.join(self.output_dir, f'model.pt.best'))
+ torch.save(state, best_ckpt)
+ logging.info(f"Update best loss: {self.val_loss_step_or_eoch[self.best_step_or_epoch]:.4f}, {best_ckpt}")
+ else:
+ logging.info(f"No improvement in loss: {self.val_loss_step_or_eoch[ckpt_name]:.4f} > {self.val_loss_step_or_eoch[self.best_step_or_epoch]:.4f}")
else:
- logging.info(f"No improvement in acc: {self.val_acc_list[self.best_acc_idx]}")
-
+ print("Undo")
+ self.saved_ckpts[ckpt_name] = getattr(self, f"val_{self.avg_keep_nbest_models_type}_step_or_eoch")[ckpt_name]
if self.keep_nbest_models > 0:
- self.saved_ckpts[ckpt_name] = self.val_acc_list[-1]
if len(self.saved_ckpts) > self.keep_nbest_models:
-
- min_key = min(self.saved_ckpts, key=self.saved_ckpts.get)
- if min_key in self.saved_ckpts:
- del self.saved_ckpts[min_key]
- filename = os.path.join(self.output_dir, min_key)
+ if self.avg_keep_nbest_models_type == "acc":
+ key = min(self.saved_ckpts, key=self.saved_ckpts.get)
+ else:
+ key = max(self.saved_ckpts, key=self.saved_ckpts.get)
+ if key in self.saved_ckpts:
+ del self.saved_ckpts[key]
+ filename = os.path.join(self.output_dir, key)
logging.info(f"Delete: {filename}")
if os.path.exists(filename):
os.remove(filename)
@@ -190,7 +207,7 @@
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']
@@ -213,9 +230,11 @@
if scaler is not None and 'scaler_state' in checkpoint:
scaler.load_state_dict(checkpoint['scaler_state'])
- self.val_acc_list = checkpoint["acc"]
- self.step_or_epoch = checkpoint["step_or_epoch"]
-
+ self.saved_ckpts = checkpoint["saved_ckpts"]
+ self.val_acc_step_or_eoch = checkpoint["val_acc_step_or_eoch"] if "val_acc_step_or_eoch" in checkpoint else {}
+ self.val_loss_step_or_eoch = checkpoint["val_loss_step_or_eoch"] if "val_loss_step_or_eoch" in checkpoint else {}
+ self.best_step_or_epoch = checkpoint["best_step_or_epoch"] if "best_step_or_epoch" in checkpoint else ""
+ model.to(self.device)
print(f"Checkpoint loaded successfully from '{ckpt}'")
else:
print(f"No checkpoint found at '{ckpt}', does not resume status!")
@@ -233,12 +252,15 @@
dataloader_val=None,
epoch=None,
writer=None,
+ **kwargs,
):
"""
Defines the training process for a single epoch with gradient accumulation.
Args:
epoch (int): The current epoch number.
"""
+ if self.use_ddp or self.use_fsdp:
+ dist.barrier()
logging.info(f"Train epoch: {epoch}, rank: {self.local_rank}\n")
model.train()
@@ -247,11 +269,12 @@
# Initialize the gradient accumulation
optim.zero_grad()
speed_stats = {}
- time5 = time.perf_counter()
+
iterator_stop = torch.tensor(0).to(self.device)
- dist.barrier()
- print(f"before iter, iterator_stop: {iterator_stop}\n")
+
dataloader_train.batch_sampler.set_epoch(epoch)
+ time_beg = time.perf_counter()
+ time5 = time_beg
for batch_idx, batch in enumerate(dataloader_train):
if self.use_ddp or self.use_fsdp:
dist.all_reduce(iterator_stop, dist.ReduceOp.SUM)
@@ -259,11 +282,13 @@
break
self.batch_total += 1
time1 = time.perf_counter()
- speed_stats["data_load"] = f"{time1-time5:0.3f}"
+ speed_stats["data_load"] = f"{time1-time_beg:0.3f}"
batch = to_device(batch, self.device)
-
- my_context = model.no_sync if batch_idx % accum_grad != 0 else nullcontext
+
+ my_context = nullcontext
+ if self.use_ddp or self.use_fsdp:
+ my_context = model.no_sync if batch_idx % accum_grad != 0 else my_context
with my_context():
time2 = time.perf_counter()
with maybe_autocast(self.use_fp16):
@@ -297,13 +322,13 @@
self.train_loss_avg = (self.train_loss_avg*batch_idx + loss.detach().cpu().item())/(batch_idx+1)
if "acc" in stats:
self.train_acc_avg = (self.train_acc_avg * batch_idx + stats["acc"].detach().cpu().item()) / (batch_idx + 1)
- # if self.use_ddp or self.use_fsdp:
- # train_loss_avg = torch.tensor(self.train_loss_avg, dtype=torch.float32).to(self.device)
- # train_acc_avg = torch.tensor(self.train_acc_avg, dtype=torch.float32).to(self.device)
- # dist.all_reduce(train_loss_avg, op=dist.ReduceOp.SUM)
- # dist.all_reduce(train_acc_avg, op=dist.ReduceOp.SUM)
- # self.train_loss_avg = train_loss_avg.detach().cpu().item() / self.world_size
- # self.train_acc_avg = train_acc_avg.detach().cpu().item() / self.world_size
+ if self.use_ddp or self.use_fsdp:
+ train_loss_avg = torch.tensor(self.train_loss_avg, dtype=torch.float32).to(self.device)
+ train_acc_avg = torch.tensor(self.train_acc_avg, dtype=torch.float32).to(self.device)
+ dist.all_reduce(train_loss_avg, op=dist.ReduceOp.SUM)
+ dist.all_reduce(train_acc_avg, op=dist.ReduceOp.SUM)
+ self.train_loss_avg = train_loss_avg.detach().cpu().item() / self.world_size
+ self.train_acc_avg = train_acc_avg.detach().cpu().item() / self.world_size
# Perform an optimizer step only after accumulating enough gradients
@@ -339,7 +364,7 @@
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,
@@ -350,6 +375,8 @@
stats=stats,
writer=writer,
tag="train",
+ data_split_i=kwargs.get("data_split_i", 0),
+ data_split_num=kwargs.get("data_split_num", 1),
)
if (batch_idx + 1) % self.validate_interval == 0:
@@ -357,12 +384,14 @@
model=model,
dataloader_val=dataloader_val,
epoch=epoch,
- writer=writer
+ writer=writer,
+ step=batch_idx+1,
)
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)
+ time_beg = time.perf_counter()
else:
if self.use_ddp or self.use_fsdp:
iterator_stop.fill_(1)
@@ -370,6 +399,7 @@
if self.use_ddp or self.use_fsdp:
dist.barrier()
+ iterator_stop = torch.tensor(0).to(self.device)
@@ -387,6 +417,8 @@
Args:
epoch (int): The current epoch number.
"""
+ if self.use_ddp or self.use_fsdp:
+ dist.barrier()
logging.info(f"Validate epoch: {epoch}, rank: {self.local_rank}\n")
model.eval()
@@ -395,13 +427,10 @@
speed_stats = {}
time5 = time.perf_counter()
iterator_stop = torch.tensor(0).to(self.device)
- dist.barrier()
- print(f"before iter, iterator_stop: {iterator_stop}\n")
+ dataloader_val.batch_sampler.set_epoch(epoch)
for batch_idx, batch in enumerate(dataloader_val):
if self.use_ddp or self.use_fsdp:
dist.all_reduce(iterator_stop, dist.ReduceOp.SUM)
- if epoch >= 1:
- print(f"iterator_stop: {iterator_stop}\n")
if iterator_stop > 0:
break
time1 = time.perf_counter()
@@ -417,7 +446,7 @@
# Apply weighted averaging for loss and stats
loss = (loss * weight.type(loss.dtype)).sum()
# if distributed, this method can also apply all_reduce()
- stats, weight = recursive_average(stats, weight, distributed=True)
+ # stats, weight = recursive_average(stats, weight, distributed=True)
if self.use_ddp or self.use_fsdp:
dist.all_reduce(weight, op=dist.ReduceOp.SUM)
# Now weight is summation over all workers
@@ -432,15 +461,15 @@
self.val_loss_avg = (self.val_loss_avg*batch_idx + loss.detach().cpu().item())/(batch_idx+1)
if "acc" in stats:
self.val_acc_avg = (self.val_acc_avg * batch_idx + stats["acc"].detach().cpu().item()) / (batch_idx + 1)
- # if self.use_ddp or self.use_fsdp:
- # val_loss_avg = torch.tensor(self.val_loss_avg, dtype=torch.float32).to(self.device)
- # val_acc_avg = torch.tensor(self.val_acc_avg, dtype=torch.float32).to(self.device)
- # dist.all_reduce(val_loss_avg, op=dist.ReduceOp.SUM)
- # 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
-
- batch_num_epoch = -1
+ if self.use_ddp or self.use_fsdp:
+ val_loss_avg = torch.tensor(self.val_loss_avg, dtype=torch.float32).to(self.device)
+ val_acc_avg = torch.tensor(self.val_acc_avg, dtype=torch.float32).to(self.device)
+ dist.all_reduce(val_loss_avg, op=dist.ReduceOp.SUM)
+ 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)
self.log(epoch, batch_idx,
@@ -457,14 +486,18 @@
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)
+
+ if kwargs.get("step", None) is None:
+ ckpt_name = f'model.pt.ep{epoch}'
+ else:
+ ckpt_name = f'model.pt.ep{epoch}.{kwargs.get("step")}'
+ self.val_acc_step_or_eoch[ckpt_name] = self.val_acc_avg
+ self.val_loss_step_or_eoch[ckpt_name] = self.val_loss_avg
model.train()
-
-
if self.use_ddp or self.use_fsdp:
dist.barrier()
+ iterator_stop = torch.tensor(0).to(self.device)
def log(self,
@@ -477,6 +510,9 @@
stats=None,
writer=None,
tag="train",
+ data_split_i=0,
+ data_split_num=1,
+ **kwargs,
):
if (batch_idx + 1) % self.log_interval == 0:
@@ -496,10 +532,11 @@
f"{tag}, "
f"rank: {self.local_rank}, "
f"epoch: {epoch}/{self.max_epoch}, "
+ f"data_slice: {data_split_i}/{data_split_num}, "
f"step: {batch_idx + 1}/{batch_num_epoch}, total step: {self.batch_total}, "
f"(loss_avg_rank: {loss:.3f}), "
f"(loss_avg_epoch: {loss_avg_epoch:.3f}), "
- f"(ppl_avg_epoch: {math.exp(loss_avg_epoch):.3f}), "
+ f"(ppl_avg_epoch: {math.exp(loss_avg_epoch):.3e}), "
f"(acc_avg_epoch: {acc_avg_epoch:.3f}), "
f"(lr: {lr:.3e}), "
f"{[(k, round(v.detach().cpu().item(), 3)) for k, v in stats.items()]}, "
--
Gitblit v1.9.1