From e32e1a7747f0eeeb68237175e0faf391d422891b Mon Sep 17 00:00:00 2001
From: 嘉渊 <wangjiaming.wjm@alibaba-inc.com>
Date: 星期四, 11 五月 2023 17:42:09 +0800
Subject: [PATCH] update repo
---
funasr/bin/train.py | 49 +++++++++++++++++++++----------------------------
1 files changed, 21 insertions(+), 28 deletions(-)
diff --git a/funasr/bin/train.py b/funasr/bin/train.py
index 3e3f598..56a9269 100755
--- a/funasr/bin/train.py
+++ b/funasr/bin/train.py
@@ -23,7 +23,6 @@
from funasr.utils.prepare_data import prepare_data
from funasr.utils.types import int_or_none
from funasr.utils.types import str2bool
-from funasr.utils.types import str2triple_str
from funasr.utils.types import str_or_none
from funasr.utils.yaml_no_alias_safe_dump import yaml_no_alias_safe_dump
@@ -161,6 +160,7 @@
)
parser.add_argument(
"--patience",
+ type=int_or_none,
default=None,
help="Number of epochs to wait without improvement "
"before stopping the training",
@@ -260,6 +260,12 @@
"training phase. If None is given, it is decided according the number "
"of training samples automatically .",
)
+ parser.add_argument(
+ "--use_tensorboard",
+ type=str2bool,
+ default=True,
+ help="Enable tensorboard logging",
+ )
# pretrained model related
parser.add_argument(
@@ -310,42 +316,24 @@
help=f"The keyword arguments for dataset",
)
parser.add_argument(
- "--train_data_file",
+ "--data_dir",
type=str,
default=None,
- help="train_list for large dataset",
+ help="root path of data",
)
parser.add_argument(
- "--valid_data_file",
+ "--train_set",
type=str,
- default=None,
- help="valid_list for large dataset",
+ default="train",
+ help="train dataset",
)
parser.add_argument(
- "--train_data_path_and_name_and_type",
- type=str2triple_str,
- action="append",
- default=[],
- help="e.g. '--train_data_path_and_name_and_type some/path/a.scp,foo,sound'. ",
- )
- parser.add_argument(
- "--valid_data_path_and_name_and_type",
- type=str2triple_str,
- action="append",
- default=[],
- )
- parser.add_argument(
- "--train_shape_file",
+ "--valid_set",
type=str,
- action="append",
- default=[],
+ default="validation",
+ help="dev dataset",
)
- parser.add_argument(
- "--valid_shape_file",
- type=str,
- action="append",
- default=[],
- )
+
parser.add_argument(
"--use_preprocessor",
type=str2bool,
@@ -514,6 +502,10 @@
prepare_data(args, distributed_option)
model = build_model(args)
+ model = model.to(
+ dtype=getattr(torch, args.train_dtype),
+ device="cuda" if args.ngpu > 0 else "cpu",
+ )
optimizers = build_optimizer(args, model=model)
schedulers = build_scheduler(args, optimizers)
@@ -521,6 +513,7 @@
distributed_option.dist_rank,
distributed_option.local_rank))
logging.info(pytorch_cudnn_version())
+ logging.info("Args: {}".format(args))
logging.info(model_summary(model))
logging.info("Optimizer: {}".format(optimizers))
logging.info("Scheduler: {}".format(schedulers))
--
Gitblit v1.9.1