From 26ab38b56cb4f69cc82e5d58907f2a57f6f2cbdd Mon Sep 17 00:00:00 2001
From: 游雁 <zhifu.gzf@alibaba-inc.com>
Date: 星期三, 21 二月 2024 14:22:14 +0800
Subject: [PATCH] update train recipe
---
examples/aishell/conformer/run.sh | 1 +
funasr/train_utils/trainer.py | 4 ++++
examples/aishell/paraformer/run.sh | 1 +
examples/aishell/branchformer/run.sh | 1 +
examples/aishell/e_branchformer/run.sh | 1 +
examples/aishell/transformer/run.sh | 1 +
6 files changed, 9 insertions(+), 0 deletions(-)
diff --git a/examples/aishell/branchformer/run.sh b/examples/aishell/branchformer/run.sh
index e6d6442..448e936 100755
--- a/examples/aishell/branchformer/run.sh
+++ b/examples/aishell/branchformer/run.sh
@@ -174,6 +174,7 @@
++output_dir="${inference_dir}/${JOB}" \
++device="${inference_device}" \
++ncpu=1 \
+ ++disable_log=true \
++batch_size="${inference_batch_size}" &> ${_logdir}/log.${JOB}.txt
}&
diff --git a/examples/aishell/conformer/run.sh b/examples/aishell/conformer/run.sh
index 1fbf1f1..f5d993a 100755
--- a/examples/aishell/conformer/run.sh
+++ b/examples/aishell/conformer/run.sh
@@ -173,6 +173,7 @@
++output_dir="${inference_dir}/${JOB}" \
++device="${inference_device}" \
++ncpu=1 \
+ ++disable_log=true \
++batch_size="${inference_batch_size}" &> ${_logdir}/log.${JOB}.txt
}&
diff --git a/examples/aishell/e_branchformer/run.sh b/examples/aishell/e_branchformer/run.sh
index 4f65ac8..ce6d346 100755
--- a/examples/aishell/e_branchformer/run.sh
+++ b/examples/aishell/e_branchformer/run.sh
@@ -174,6 +174,7 @@
++output_dir="${inference_dir}/${JOB}" \
++device="${inference_device}" \
++ncpu=1 \
+ ++disable_log=true \
++batch_size="${inference_batch_size}" &> ${_logdir}/log.${JOB}.txt
}&
diff --git a/examples/aishell/paraformer/run.sh b/examples/aishell/paraformer/run.sh
index bc03fb3..f96b192 100755
--- a/examples/aishell/paraformer/run.sh
+++ b/examples/aishell/paraformer/run.sh
@@ -173,6 +173,7 @@
++output_dir="${inference_dir}/${JOB}" \
++device="${inference_device}" \
++ncpu=1 \
+ ++disable_log=true \
++batch_size="${inference_batch_size}" &> ${_logdir}/log.${JOB}.txt
}&
diff --git a/examples/aishell/transformer/run.sh b/examples/aishell/transformer/run.sh
index 12aef53..c282d7f 100755
--- a/examples/aishell/transformer/run.sh
+++ b/examples/aishell/transformer/run.sh
@@ -174,6 +174,7 @@
++output_dir="${inference_dir}/${JOB}" \
++device="${inference_device}" \
++ncpu=1 \
+ ++disable_log=true \
++batch_size="${inference_batch_size}" &> ${_logdir}/log.${JOB}.txt
}&
diff --git a/funasr/train_utils/trainer.py b/funasr/train_utils/trainer.py
index 61b9004..f4ef773 100644
--- a/funasr/train_utils/trainer.py
+++ b/funasr/train_utils/trainer.py
@@ -3,6 +3,7 @@
import torch
import logging
from tqdm import tqdm
+from datetime import datetime
import torch.distributed as dist
from contextlib import nullcontext
# from torch.utils.tensorboard import SummaryWriter
@@ -283,7 +284,10 @@
torch.cuda.max_memory_reserved()/1024/1024/1024,
)
lr = self.scheduler.get_last_lr()[0]
+ time_now = datetime.now()
+ time_now = now.strftime("%Y-%m-%d %H:%M:%S")
description = (
+ f"{time_now}, "
f"rank: {self.local_rank}, "
f"epoch: {epoch}/{self.max_epoch}, "
f"step: {batch_idx+1}/{len(self.dataloader_train)}, total: {self.batch_total}, "
--
Gitblit v1.9.1