From 33d3d2084403fd34b79c835d2f2fe04f6cd8f738 Mon Sep 17 00:00:00 2001
From: 游雁 <zhifu.gzf@alibaba-inc.com>
Date: 星期三, 13 九月 2023 09:33:54 +0800
Subject: [PATCH] Merge branch 'main' of github.com:alibaba-damo-academy/FunASR add
---
egs/librispeech_100h/conformer/run.sh | 67 ++++++++++++++++++++-------------
1 files changed, 41 insertions(+), 26 deletions(-)
diff --git a/egs/librispeech_100h/conformer/run.sh b/egs/librispeech_100h/conformer/run.sh
index 354610c..41df5a4 100755
--- a/egs/librispeech_100h/conformer/run.sh
+++ b/egs/librispeech_100h/conformer/run.sh
@@ -19,8 +19,9 @@
token_type=bpe
type=sound
scp=wav.scp
-stage=1
-stop_stage=1
+speed_perturb="0.9 1.0 1.1"
+stage=0
+stop_stage=5
# feature configuration
feats_dim=80
@@ -52,9 +53,10 @@
asr_config=conf/train_asr_conformer.yaml
model_dir="baseline_$(basename "${asr_config}" .yaml)_${lang}_${token_type}_${tag}"
-inference_config=conf/decode_asr_transformer.yaml
-#inference_config=conf/decode_asr_transformer_beam60_ctc0.3.yaml
-inference_asr_model=valid.acc.ave_10best.pth
+#inference_config=conf/decode_asr_transformer_ctc0.3_beam1.yaml
+inference_config=conf/decode_asr_transformer_ctc0.3_beam5.yaml
+#inference_config=conf/decode_asr_transformer_ctc0.3_beam20.yaml
+inference_asr_model=valid.acc.ave_10best.pb
# you can set gpu num for decoding here
gpuid_list=$CUDA_VISIBLE_DEVICES # set gpus for decoding, the same as training stage by default
@@ -82,34 +84,44 @@
for x in dev-clean dev-other test-clean test-other train-clean-100; do
local/data_prep.sh ${raw_data}/LibriSpeech/${x} ${feats_dir}/data/${x//-/_}
done
+ mkdir $feats_dir/data/$valid_set
+ dev_sets="dev_clean dev_other"
+ for file in wav.scp text; do
+ ( for f in $dev_sets; do cat $feats_dir/data/$f/$file; done ) | sort -k1 > $feats_dir/data/$valid_set/$file || exit 1;
+ done
fi
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
echo "stage 1: Feature and CMVN Generation"
- utils/compute_cmvn.sh --cmd "$train_cmd" --nj $nj --feats_dim ${feats_dim} ${feats_dir}/data/${train_set}
+ utils/compute_cmvn.sh --fbankdir ${feats_dir}/data/${train_set} --cmd "$train_cmd" --nj $nj --feats_dim ${feats_dim} --config_file "$asr_config" --scale 1.0
fi
-dict=${feats_dir}/data/lang_char/${train_set}_${bpemode}${nbpe}_units.txt
+token_list=${feats_dir}/data/lang_char/${train_set}_${bpemode}${nbpe}_units.txt
bpemodel=${feats_dir}/data/lang_char/${train_set}_${bpemode}${nbpe}
-echo "dictionary: ${dict}"
+echo "dictionary: ${token_list}"
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
### Task dependent. You have to check non-linguistic symbols used in the corpus.
echo "stage 2: Dictionary and Json Data Preparation"
mkdir -p ${feats_dir}/data/lang_char/
- echo "<blank>" > ${dict}
- echo "<s>" >> ${dict}
- echo "</s>" >> ${dict}
+ echo "<blank>" > ${token_list}
+ echo "<s>" >> ${token_list}
+ echo "</s>" >> ${token_list}
cut -f 2- -d" " ${feats_dir}/data/${train_set}/text > ${feats_dir}/data/lang_char/input.txt
- spm_train --input=${feats_dir}/data/lang_char/input.txt --vocab_size=${nbpe} --model_type=${bpemode} --model_prefix=${bpemodel} --input_sentence_size=100000000
- spm_encode --model=${bpemodel}.model --output_format=piece < ${feats_dir}/data/lang_char/input.txt | tr ' ' '\n' | sort | uniq | awk '{print $0}' >> ${dict}
- echo "<unk>" >> ${dict}
+ local/spm_train.py --input=${feats_dir}/data/lang_char/input.txt --vocab_size=${nbpe} --model_type=${bpemode} --model_prefix=${bpemodel} --input_sentence_size=100000000
+ local/spm_encode.py --model=${bpemodel}.model --output_format=piece < ${feats_dir}/data/lang_char/input.txt | tr ' ' '\n' | sort | uniq | awk '{print $0}' >> ${token_list}
+ echo "<unk>" >> ${token_list}
fi
-
-# Training Stage
+# LM Training Stage
world_size=$gpu_num # run on one machine
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
- echo "stage 3: Training"
+ echo "stage 3: LM Training"
+fi
+
+# ASR Training Stage
+world_size=$gpu_num # run on one machine
+if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
+ echo "stage 4: ASR Training"
mkdir -p ${exp_dir}/exp/${model_dir}
mkdir -p ${exp_dir}/exp/${model_dir}/log
INIT_FILE=${exp_dir}/exp/${model_dir}/ddp_init
@@ -123,20 +135,22 @@
rank=$i
local_rank=$i
gpu_id=$(echo $CUDA_VISIBLE_DEVICES | cut -d',' -f$[$i+1])
- asr_train.py \
+ train.py \
+ --task_name asr \
--gpu_id $gpu_id \
--use_preprocessor true \
--split_with_space false \
--bpemodel ${bpemodel}.model \
--token_type $token_type \
- --dataset_type $dataset_type \
- --token_list $dict \
- --train_data_file $feats_dir/$dumpdir/${train_set}/ark_txt.scp \
- --valid_data_file $feats_dir/$dumpdir/${valid_set}/ark_txt.scp \
+ --token_list $token_list \
+ --data_dir ${feats_dir}/data \
+ --train_set ${train_set} \
+ --valid_set ${valid_set} \
+ --cmvn_file ${feats_dir}/data/${train_set}/cmvn/am.mvn \
+ --speed_perturb ${speed_perturb} \
--resume true \
--output_dir ${exp_dir}/exp/${model_dir} \
--config $asr_config \
- --input_size $feats_dim \
--ngpu $gpu_num \
--num_worker_count $count \
--multiprocessing_distributed true \
@@ -150,8 +164,8 @@
fi
# Testing Stage
-if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
- echo "stage 4: Inference"
+if [ ${stage} -le 5 ] && [ ${stop_stage} -ge 5 ]; then
+ echo "stage 5: Inference"
for dset in ${test_sets}; do
asr_exp=${exp_dir}/exp/${model_dir}
inference_tag="$(basename "${inference_config}" .yaml)"
@@ -162,7 +176,7 @@
exit 0
fi
mkdir -p "${_logdir}"
- _data="${feats_dir}/${dumpdir}/${dset}"
+ _data="${feats_dir}/data/${dset}"
key_file=${_data}/${scp}
num_scp_file="$(<${key_file} wc -l)"
_nj=$([ $inference_nj -le $num_scp_file ] && echo "$inference_nj" || echo "$num_scp_file")
@@ -183,6 +197,7 @@
--njob ${njob} \
--gpuid_list ${gpuid_list} \
--data_path_and_name_and_type "${_data}/${scp},speech,${type}" \
+ --cmvn_file ${feats_dir}/data/${train_set}/cmvn/am.mvn \
--key_file "${_logdir}"/keys.JOB.scp \
--asr_train_config "${asr_exp}"/config.yaml \
--asr_model_file "${asr_exp}"/"${inference_asr_model}" \
--
Gitblit v1.9.1