感谢,在tiny目录下sh run_train.sh时报错
Created by: Kangzhiwen
按照readme一步步来的,然后遇到了swig_decoders不存在的问题,按照issue解决后,继续执行train报如下错误: W0409 11:01:41.920228 2557 init.cc:209] Warning: PaddlePaddle catches a failure signal, it may not work properly W0409 11:01:41.920295 2557 init.cc:211] You could check whether you killed PaddlePaddle thread/process accidentally or report the case to PaddlePaddle W0409 11:01:41.920307 2557 init.cc:214] The detail failure signal is:
W0409 11:01:41.920320 2557 init.cc:217] *** Aborted at 1586401301 (unix time) try "date -d @1586401301" if you are using GNU date *** W0409 11:01:41.923130 2557 init.cc:217] PC: @ 0x0 (unknown) run_train.sh: 行 35: 2557 段错误 (吐核)CUDA_VISIBLE_DEVICES=0,1,2,3 python -u train.py --batch_size=4 --num_epoch=20 --num_conv_layers=2 --num_rnn_layers=3 --rnn_layer_size=2048 --num_iter_print=1 --save_epoch=1 --num_samples=64 --learning_rate=1e-5 --max_duration=27.0 --min_duration=0.0 --test_off=False --use_sortagrad=True --use_gru=False --use_gpu=False --is_local=True --share_rnn_weights=True --train_manifest='data/tiny/manifest.tiny' --dev_manifest='data/tiny/manifest.tiny' --mean_std_path='data/tiny/mean_std.npz' --vocab_path='data/tiny/vocab.txt' --output_model_dir='./checkpoints/tiny' --augment_conf_path='conf/augmentation.config' --specgram_type='linear' --shuffle_method='batch_shuffle_clipped' Failed in training!
paddlepaddle安装如下,安装的无GPU版本,非docker安装: PaddlePaddle 1.7.1, compiled with with_avx: ON with_gpu: OFF with_mkl: ON with_mkldnn: ON with_python: ON
DeepSpeech也是非Docker安装。
现在解决不了这个问题,请大家指教。