From 35759e53af7cb5cc96489bc00dd44884bb9423c2 Mon Sep 17 00:00:00 2001
From: tangyunxiang <546783735@qq.com>
Date: Wed, 12 Jul 2023 11:02:01 +0800
Subject: [PATCH 1/2] =?UTF-8?q?=E6=96=B0=E5=A2=9EEspNet2=E5=9C=A8=E7=BA=BF?=
 =?UTF-8?q?=E6=8E=A8=E7=90=86=E6=8C=87=E5=AF=BC?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 PyTorch/built-in/audio/ESPnet2_for_PyTorch/README.md | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/PyTorch/built-in/audio/ESPnet2_for_PyTorch/README.md b/PyTorch/built-in/audio/ESPnet2_for_PyTorch/README.md
index d902084258..9ccca8c254 100644
--- a/PyTorch/built-in/audio/ESPnet2_for_PyTorch/README.md
+++ b/PyTorch/built-in/audio/ESPnet2_for_PyTorch/README.md
@@ -152,6 +152,14 @@ ESPNet是一套基于E2E的开源工具包，可进行语音识别等任务。
    
    --fp32开启FP32模式
 
+   - 在线推理
+
+     启动在线推理
+   
+     ```
+     bash ./test/train_eval_1p.sh
+     ```
+
 3. 启动训练后，日志输出路径为：<espnet-root>/egs2/aishell/asr1/nohup.out ，该日志中会打印二级日志（各个stage日志）的相对路径。
 如：stage 11 的日志路径为：“exp/asr_train_asr_conformer_raw_zh_char_sp/train.log”
 
-- 
Gitee


From 2fc4137267c3168f5b569631fba128a2cfd978ff Mon Sep 17 00:00:00 2001
From: tangyunxiang <546783735@qq.com>
Date: Wed, 12 Jul 2023 11:02:33 +0800
Subject: [PATCH 2/2] =?UTF-8?q?=E6=96=B0=E5=A2=9EEspNet2=E5=9C=A8=E7=BA=BF?=
 =?UTF-8?q?=E6=8E=A8=E7=90=86=E8=84=9A=E6=9C=AC?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 .../ESPnet2_for_PyTorch/test/train_eval_1p.sh | 119 ++++++++++++++++++
 1 file changed, 119 insertions(+)
 create mode 100644 PyTorch/built-in/audio/ESPnet2_for_PyTorch/test/train_eval_1p.sh

diff --git a/PyTorch/built-in/audio/ESPnet2_for_PyTorch/test/train_eval_1p.sh b/PyTorch/built-in/audio/ESPnet2_for_PyTorch/test/train_eval_1p.sh
new file mode 100644
index 0000000000..32857db6fe
--- /dev/null
+++ b/PyTorch/built-in/audio/ESPnet2_for_PyTorch/test/train_eval_1p.sh
@@ -0,0 +1,119 @@
+#!/bin/bash
+source ./test/env_npu.sh
+
+#当前路径,不需要修改
+cur_path=`pwd`
+cur_path_last_dirname=${cur_path##*/}
+if [ x"${cur_path_last_dirname}" == x"test" ];then
+    test_path_dir=${cur_path}
+    cd ..
+    cur_path=`pwd`
+else
+    test_path_dir=${cur_path}/test
+fi
+
+#集合通信参数,不需要修改
+export RANK_SIZE=1
+
+
+#网络名称,同目录名称,需要模型审视修改
+Network="ESPnet2_for_PyTorch"
+
+#混合精度
+use_amp=true
+
+# 指定训练所使用的npu device卡id, 暂不支持修改
+device_id=0
+
+for para in $*
+do
+    if [[ $para == --use_amp* ]];then
+        use_amp=`echo ${para#*=}`
+    fi
+    if [[ $para == --fp32 ]];then
+        fp32=`echo ${para#*=}`
+        export ALLOW_FP32=True
+    fi
+done
+
+export USE_AMP=${use_amp}
+
+# 校验是否指定了device_id,分动态分配device_id与手动指定device_id,此处不需要修改
+if [ ${device_id} ];then
+    export ASCEND_DEVICE_ID=${device_id}
+    echo "device id is ${ASCEND_DEVICE_ID}"
+else
+    "[Error] device id must be config"
+    exit 1
+fi
+
+#创建DeviceID输出目录，不需要修改
+if [ -d $test_path_dir/output/${ASCEND_DEVICE_ID} ];then
+    rm -rf $test_path_dir/output/$ASCEND_DEVICE_ID
+    mkdir -p $test_path_dir/output/$ASCEND_DEVICE_ID
+else
+    mkdir -p $test_path_dir/output/$ASCEND_DEVICE_ID
+fi
+
+#################启动训练脚本#################
+
+# 必要参数替换配置文件
+cd $cur_path/egs2/aishell/asr1
+
+# 修改配置文件超参
+conf_file=$cur_path/egs2/aishell/asr1/conf/tuning/train_asr_conformer.yaml
+lm_conf_file=$cur_path/egs2/aishell/asr1/conf/tuning/train_lm_transformer.yaml
+ori_epoch=`cat $conf_file | grep max_epoch:`
+ori_batch_bins=`cat $conf_file | grep batch_bins:`
+ori_lr=`cat $conf_file | grep lr:`
+ori_use_amp=`cat $conf_file | grep use_amp:`
+lm_ori_use_amp=`cat $lm_conf_file | grep use_amp:`
+sed -i "s|$ori_epoch|max_epoch: 50|g" $conf_file
+sed -i "s|$ori_batch_bins|batch_bins: 4000000|g" $conf_file
+sed -i "s|$ori_lr|   lr: 0.0005|g" $conf_file
+sed -i "s|$ori_use_amp|use_amp: $use_amp|g" $conf_file
+sed -i "s|$lm_ori_use_amp|use_amp: $use_amp|g" $lm_conf_file
+start_time=$(date +%s)
+
+nohup bash run.sh \
+  --stage 12 \
+  --use_lm true \
+  --ngpu 1 &
+wait
+
+result=$cur_path/egs2/aishell/asr1/exp/asr_train_asr_conformer_raw_zh_char_sp/RESULTS.md
+
+##################获取推理数据################
+#推理结束时间，不需要修改
+end_time=$(date +%s)
+e2e_time=$(( $end_time - $start_time ))
+
+#结果打印，不需要修改
+echo "------------------ Final result ------------------"
+#输出性能FPS，需要模型审视修改
+# 计算公式为：数据集数量 / 训练总时间
+FPS=`awk 'BEGIN{printf "%.2f",(104765 / '$TOTAL_TIME')}'`
+
+#输出训练精度,需要模型审视修改
+dev_accuracy=`grep "valid.acc.ave/dev" ${result} | tail -n 1 | awk -F "|" '{print$5}'`
+test_accuracy=`grep "valid.acc.ave/test" ${result} | tail -n 1 | awk -F "|" '{print$5}'`
+
+#打印，不需要修改
+echo "Final Performance waves/sec : $FPS"
+
+#打印，不需要修改
+echo "Final Validate Accuracy : ${test_accuracy}"
+echo "E2E Validating Duration sec : $e2e_time"
+
+#获取性能数据，不需要修改
+#吞吐量
+ActualFPS=${FPS}
+
+#关键信息打印到${CaseName}.log中，不需要修改
+echo "Network = ${Network}" > ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "RankSize = ${RANK_SIZE}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "DeviceType = ${DeviceType}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "CaseName = ${CaseName}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "ActualFPS = ${ActualFPS}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "ValidateAccuracy = ${test_accuracy}" >> $test_path_dir/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "E2EValidatingTime = ${e2e_time}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
\ No newline at end of file
-- 
Gitee