加入 Gitee
与超过 1200万 开发者一起发现、参与优秀开源项目,私有仓库也完全免费 :)
免费加入
文件
克隆/下载
.compute 778 Bytes
一键复制 编辑 原始数据 按行查看 历史
#!/bin/bash
source ../tmp/venv/bin/activate
data="${DATA_ROOT}/shared/data"
fis="${data}/LDC/fisher"
swb="${data}/LDC/LDC97S62/swb"
lbs="${data}/OpenSLR/LibriSpeech/librivox"
python3 -u DeepSpeech.py \
--train_files "${fis}-train.csv","${swb}-train.csv","${lbs}-train-clean-100.csv","${lbs}-train-clean-360.csv","${lbs}-train-other-500.csv" \
--dev_files "${lbs}-dev-clean.csv"\
--test_files "${lbs}-test-clean.csv" \
--train_batch_size 24 \
--dev_batch_size 48 \
--test_batch_size 48 \
--n_hidden 2048 \
--learning_rate 0.0001 \
--dropout_rate 0.2 \
--epoch 13 \
--display_step 0 \
--validation_step 1 \
--checkpoint_dir "../keep" \
--summary_dir "../keep/summaries" \
--decoder_library_path "../tmp/native_client/libctc_decoder_with_kenlm.so"
Loading...
马建仓 AI 助手
尝试更多
代码解读
代码找茬
代码优化