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/aishell2/transformerLM/run.sh | 27 +++++----------------------
1 files changed, 5 insertions(+), 22 deletions(-)
diff --git a/egs/aishell2/transformerLM/run.sh b/egs/aishell2/transformerLM/run.sh
index 28e3762..b04c643 100755
--- a/egs/aishell2/transformerLM/run.sh
+++ b/egs/aishell2/transformerLM/run.sh
@@ -34,7 +34,7 @@
tag=exp1
model_dir="baseline_$(basename "${lm_config}" .yaml)_${lang}_${token_type}_${tag}"
lm_exp=${exp_dir}/exp/${model_dir}
-inference_lm=valid.loss.ave.pth # Language model path for decoding.
+inference_lm=valid.loss.ave.pb # Language model path for decoding.
stage=0
stop_stage=3
@@ -92,25 +92,6 @@
echo "Error: not supported --token_type '${token_type}'"
exit 2
fi
-
- ## use_word_lm=false
- ## # Create word-list for word-LM training
- ## if ${use_word_lm} && [ "${token_type}" != word ]; then
- ## echo "Generate word level token_list from ${lm_train_text}"
- ## python -m funasr.bin.tokenize_text \
- ## --token_type word \
- ## --input "${lm_train_text}" \
- ## --output "${token_list}" \
- ## --field 2- \
- ## --cleaner "${cleaner}" \
- ## --g2p "${g2p}" \
- ## --write_vocabulary true \
- ## --vocabulary_size "${word_vocab_size}" \
- ## --add_symbol "${blank}:0" \
- ## --add_symbol "${sos}:1" \
- ## --add_symbol "${eos}:2" \
- ## --add_symbol "${oov}:-1"
- ## fi
lm_token_list="${token_list}"
@@ -232,14 +213,16 @@
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
echo "Stage 3: Calc perplexity: ${lm_test_text}"
- python ../../../funasr/bin/lm_inference.py \
- --output_dir "${lm_exp}/perplexity_test" \
+ python ../../../funasr/bin/lm_inference_launch.py \
+ --output_dir "${lm_exp}/perplexity_test/output.1" \
--ngpu "${gpu_num}" \
--batch_size 1 \
--train_config "${lm_exp}"/config.yaml \
--model_file "${lm_exp}/${inference_lm}" \
--data_path_and_name_and_type "${lm_test_text},text,text" \
--num_workers 1 \
+ --gpuid_list 0 \
+ --mode "transformer" \
--split_with_space false
fi
--
Gitblit v1.9.1