speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

fix bug

Browse Source
This commit is contained in:
游雁 2024-06-10 10:10:47 +08:00
parent 1cd2812d46
commit 5ca83889c2
1 changed files with 18 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
examples/industrial_data_pretraining/llm_asr/demo_speech2text.sh
View File

@ -1,27 +1,28 @@
ckpt_dir="/nfs/beinian.lzr/workspace/GPT-4o/Exp/exp6/5m-8gpu/exp6_speech2text_linear_ddp_0609"
ckpt_id="model.pt.ep0.90000"
jsonl_dir="/nfs/beinian.lzr/workspace/GPT-4o/Data/Speech2Text/TestData"
out_dir="${ckpt_dir}/inference-${ckpt_id}"
mkdir -p ${out_dir}
device="cuda:0"
ckpt_id=$1
device=$2
#for data_set in "librispeech_test_clean_speech2text.jsonl" "librispeech_test_other_speech2text.jsonl"; do
# jsonl=${jsonl_dir}/${data_set}
# output_dir=${out_dir}/${data_set}
# mkdir -p ${output_dir}
# pred_file=${output_dir}/1best_recog/text_tn
# ref_file=${output_dir}/1best_recog/label
#
# python ./demo_speech2text.py ${ckpt_dir} ${ckpt_id} ${jsonl} ${output_dir} ${device}
#
# python /mnt/workspace/zhifu.gzf/codebase/FunASR/funasr/metrics/wer.py ++ref_file=${ref_file} ++hyp_file=${pred_file} ++cer_file=${pred_file}.cer ++cn_postprocess=false
#
#done
ckpt_dir="/nfs/beinian.lzr/workspace/GPT-4o/Exp/exp6/5m-8gpu/exp6_speech2text_linear_ddp_0609"
jsonl_dir="/nfs/beinian.lzr/workspace/GPT-4o/Data/Speech2Text/TestData"
out_dir="${ckpt_dir}/inference-${ckpt_id}"
mkdir -p ${out_dir}
for data_set in "librispeech_test_clean_speech2text.jsonl" "librispeech_test_other_speech2text.jsonl"; do
jsonl=${jsonl_dir}/${data_set}
output_dir=${out_dir}/${data_set}
mkdir -p ${output_dir}
pred_file=${output_dir}/1best_recog/text_tn
ref_file=${output_dir}/1best_recog/label
python ./demo_speech2text.py ${ckpt_dir} ${ckpt_id} ${jsonl} ${output_dir} ${device}
python /mnt/workspace/zhifu.gzf/codebase/FunASR/funasr/metrics/wer.py ++ref_file=${ref_file} ++hyp_file=${pred_file} ++cer_file=${pred_file}.cer ++cn_postprocess=false
done
for data_set in "aishell1_test_speech2text.jsonl" "aishell2_ios_test_speech2text.jsonl" "librispeech_test_other_speech2text.jsonl"; do