Apologies for the very basic question - but for the life of me, I cannot get a trained model exported. I use the following command:
sudo python3 ./DeepSpeech.py --train_files ../data/Voice_180207/train.csv --dev_files ../data/Voice_180207/train.csv --test_files ../data/Voice_180207/train.csv --export-dir ../models/Voice_180207 --epoch 1 --checkpoint_dir ../tmp_checkpoints
Everything works fine, except the “Models exported at” statement never shows up (left out the “src” contents for brevity):
Preprocessing ['../data/Voice_180207/train.csv']
Preprocessing done
Preprocessing ['../data/Voice_180207/train.csv']
Preprocessing done
W Parameter --validation_step needs to be >0 for early stopping to work
I STARTING Optimization
I Training epoch 0...
I Training of Epoch 0 - loss: 15559.583984
I FINISHED Optimization - training time: 0:17:52
100% (1 of 1) |#####################################| Elapsed Time: 0:00:00 Time: 0:00:00
Preprocessing ['../data/Voice_180207/train.csv']
Preprocessing done
Computing acoustic model predictions...
100% (1 of 1) |#####################################| Elapsed Time: 0:01:01 Time: 0:01:01
Decoding predictions...
100% (1 of 1) |#####################################| Elapsed Time: 0:00:07 Time: 0:00:07
Test - WER: 1.000000, CER: 1266.000000, loss: 30125.867188
--------------------------------------------------------------------------------
WER: 1.000000, CER: 1266.000000, loss: 30125.867188
- src: "..."
- res: "i "