| | |
|---|
| | | data_dir= |
|---|
| | | exp_dir= |
|---|
| | | model_name=speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch |
|---|
| | | model_revision="v1.0.3" # please do not modify the model revision |
|---|
| | | model_revision="v1.0.4" # please do not modify the model revision |
|---|
| | | inference_nj=32 |
|---|
| | | gpuid_list="0,1" # set gpus, e.g., gpuid_list="0,1" |
|---|
| | | gpuid_list="0" # set gpus, e.g., gpuid_list="0,1" |
|---|
| | | ngpu=$(echo $gpuid_list | awk -F "," '{print NF}') |
|---|
| | | njob=4 # the number of jobs for each gpu |
|---|
| | | njob=1 # the number of jobs for each gpu |
|---|
| | | gpu_inference=true # Whether to perform gpu decoding, set false for cpu decoding |
|---|
| | | |
|---|
| | | if ${gpu_inference}; then |
|---|
