You can obtain the TensorFlow version with
python -c "import tensorflow as tf; print(tf.GIT_VERSION, tf.VERSION)"
Trying to train DeepSpeech on the Arabic common voice Datasets using google Colaboratory, and after cloning the latest version of DeepSpeech and following the exact steps on the documentation, I ran into this error after trying to start the training (or lance the --helpfull )
Traceback (most recent call last):
File "DeepSpeech.py", line 7, in
from deepspeech_training import train as ds_train
File "/content/drive/My Drive/DeepSpeech/training/deepspeech_training/train.py", line 30, in
from .evaluate import evaluate
File "/content/drive/My Drive/DeepSpeech/training/deepspeech_training/evaluate.py", line 26, in
check_ctcdecoder_version()
File "/content/drive/My Drive/DeepSpeech/training/deepspeech_training/util/helpers.py", line 53, in check_ctcdecoder_version
rv = semver.compare(ds_version_s, decoder_version_s)
File "/usr/local/lib/python3.6/dist-packages/semver.py", line 452, in compare
v1, v2 = parse(ver1), parse(ver2)
File "/usr/local/lib/python3.6/dist-packages/semver.py", line 74, in parse
raise ValueError("%s is not valid SemVer string" % version)
ValueError: ../../VERSION is not valid SemVer string
Are you sure you properly git clone? Can you hand verify that the root VERSION file is indeed here?
can you check using path without spaces just in case?
I think it's Google Colab somehow. That string is what's shown in git commits for the symlink, but on clone it should turn into a real symlink.
can you check using path without spaces just in case?
I re-cloned the repo in the /content directory to avoid spaces in path and it solved the error, thank you.
I think it's Google Colab somehow. That string is what's shown in git commits for the symlink, but on clone it should turn into a real symlink.
Yes its colab problem, they use spaces in their drive repo which mess with the path in files, thank you.