examples/aishell/paraformer/run.sh
@@ -15,7 +15,7 @@ nj=32 inference_device="cuda" #"cpu" inference_checkpoint="model.pt" inference_checkpoint="model.pt.avg10" inference_scp="wav.scp" inference_batch_size=32 @@ -76,7 +76,7 @@ --config-path "${workspace}/conf" \ --config-name "${config}" \ ++train_data_set_list="${feats_dir}/data/${train_set}/audio_datasets.jsonl" \ ++cmvn_file="${feats_dir}/data/${train_set}/cmvn.json" \ ++cmvn_file="${feats_dir}/data/${train_set}/cmvn.json" fi token_list=${feats_dir}/data/${lang}_token_list/$token_type/tokens.txt
