Trie file version mismatch 4 instead of expected 3 update your trie file

[JRMeyer]

>>> javi.rahman
[August 18, 2019, 7:37pm]

Hi Deepspeech,

16.04 LTS
NO

from the 'pip list' command I can see the below installed version of
both. Could you please correct me. slash
deepspeech 0.6.0a4 slash
ds-ctcdecoder 0.5.1

I am getting slash -- slash > Error: Trie file version mismatch (4 instead of
expected 3). Update your trie file.

Here is the output of the command

I Saved new best validating model with loss 149.557864 to:
/home/jargon-javi/speech/tools/backup/checkout/best_dev-120 slash
Epoch 1 slash | Training slash | Elapsed Time: 0:00:06 slash | Steps: 5 slash | Loss:
175.983118 slash
Epoch 1 slash | Validation slash | Elapsed Time: 0:00:00 slash | Steps: 3 slash | Loss:
150.642143 slash | Dataset:
/home/jargon-javi/speech/tools/backup/train/dev.csv slash
Epoch 2 slash | Training slash | Elapsed Time: 0:00:06 slash | Steps: 5 slash | Loss:
174.152228 slash
Epoch 2 slash | Validation slash | Elapsed Time: 0:00:00 slash | Steps: 3 slash | Loss:
150.533025 slash | Dataset:
/home/jargon-javi/speech/tools/backup/train/dev.csv slash
Epoch 3 slash | Training slash | Elapsed Time: 0:00:06 slash | Steps: 5 slash | Loss:
172.228995 slash
Epoch 3 slash | Validation slash | Elapsed Time: 0:00:00 slash | Steps: 3 slash | Loss:
151.794245 slash | Dataset:
/home/jargon-javi/speech/tools/backup/train/dev.csv slash
I Early stop triggered as (for last 4 steps) validation loss: 151.794245
with standard deviation: 0.487455 and mean: 150.244344 slash
I FINISHED optimization in 0:00:26.650283 slash
Error: Trie file version mismatch (4 instead of expected 3). Update your
trie file. slash
terminate called after throwing an instance of 'int' slash
Aborted (core dumped)

[This is an archived TTS discussion thread from discourse.mozilla.org/t/trie-file-version-mismatch-4-instead-of-expected-3-update-your-trie-file]

[JRMeyer]

>>> lissyx
[August 19, 2019, 7:24am]



> deepspeech 0.6.0a4 slash
> ds-ctcdecoder 0.5.1

Ok, can you explain what you are doing ? Because this already looks like
your issue, mixing incompatible versions.

[Archived Post]

[JRMeyer]

>>> javi.rahman
[August 19, 2019, 7:32am]

,

I am running ./Deepspeech.py commands and exporting the model.

python -u DeepSpeech.py slash
/home/javi/speech/tools/DeepSpeechWorking/data/alphabet.txt slash
/home/javi/speech/tools/DeepSpeechWorking/data/lm.binary slash
' slash $'

[Archived Post]

[JRMeyer]

>>> lissyx
[August 19, 2019, 7:38am]



> I am running ./Deepspeech.py commands and exporting the model.

Again, you need to be more explicit in your setup. You mention
TensorFlow 1.13 and ds-ctcdecoder 0.5.1. Are you training with
DeepSpeech checkout of v0.5.1 ?

Where does your language model and trie comes from ? All of that is
inconsistent, and you don't document your exact setup.

[Archived Post]

[JRMeyer]

>>> javi.rahman
[August 19, 2019, 7:43am]



> Are you training with DeepSpeech checkout of v0.5.1 ?

yes



> Where does your language model and trie comes from ? All of that is
> inconsistent, and you don't document your exact setup.

I am using Kenlm and checking this link to slash -- slash > TUTORIAL : How I
trained a specific french model to control my
robot

{.site-icon width='32'

### kpu/kenlm

KenLM: Faster and Smaller Language Model Queries. Contribute to
kpu/kenlm development by creating an account on GitHub.

to create lm.binary and trie.

Could you please tell me which version is compatible for this slash -- slash >
deepspeech 0.6.0a4 slash
ds-ctcdecoder 0.5.1

[Archived Post]

[JRMeyer]

>>> lissyx
[August 19, 2019, 8:14am]



> I am using Kenlm and checking this link to slash -- slash > TUTORIAL : How I
> trained a specific french model to control my
> robot

That tutorial is a bit old, that's why you need to explicitely document
what you do ...



> Could you please tell me which version is compatible for this slash -- slash >
> deepspeech 0.6.0a4 slash
> ds-ctcdecoder 0.5.1

Could you please articulate a question here ? I've already explained you
that those versions can't be compatible.

[Archived Post]

[JRMeyer]

>>> lissyx
[August 19, 2019, 8:16am]

So again, we do
document everything on training and building a LM in the repo. Check
README.md and data/lm/README.md ...

[Archived Post]