@lissyx Hi bro,
Created a model for Hindi,
after training my data, at the test steps i get an error
I Restored variables from best validation checkpoint at hindi_checkpoint/best_dev-90, step 90 Testing model on data/test/test.csv Test epoch | Steps: 0 | Elapsed Time: 0:00:00 Fatal Python error: Segmentation fault
What could be the problem?
Without more context, it’s going to be hard … How have you setup things ?
I had a similar crash, resolved by re-creating the virtualenv from scratch …
Well, recorded few audios of mine and prepared the datasets.
step 1: prepared datasets, vocabs, alphabet.txt created arpa and then lm.bianry with hindi vocabs using kenlm
step 2: using native client bazel build created trie file using lm.binary
step 3: cloned deepspeech and gitcheckout 0.5.1
placed my trie and binary into data/lm/
running deepspeech.py , but after training i am getting segment error at testing steps:
is it because of low data in test?
Unlikely. You have not documented anything on how you did setup virtualenv … Did you read my reply ?
Okay ll try again creating venv from scratch.
Started from scratch creating venv. again same error
Well, sorry, but with so much details, I don’t even can try to reproduce …
@cryptoaimdy Seriously, I would like to help you, but you keep continuously not sharing your complete STR. This is making both of us loose valuable time. So once again, share detailed and complete STR of everything you do to reproduce the issue. And try to reproduce with our default data (LDC93S1, english model and LM and alphabet, our native client build), to make sure this is not something from there.
There are 10+ variables here in play, we can’t do divination from a single segfault.
Solved it today morning. Model is ready and giving loss rate like 60 average. Now creating more hindi datasets.
Do you care sharing what was the solution ?
As you suggested i started from scratch setting up venv.
According to me i think the problem was a version mismatch(binary version and DeepSpeech version). because earlier i created the lm binary and trie without using virtual env and deepspeech i was running using venv. so looks like a version mismatch.
Right, thanks, at least it confirms my first assumption. Glad to see it is working now.
Yes, setup is all what we need to do carefully.
But, for hindi my src " " part is not accurate. while testing the src is having sentences other than my original test.csv, the sentence in src is not making even sence. its kind of ‘abcsdksbfdfa’(hindi language abcd)
WER: 0.000000, CER: 0.000000, loss: 7.111163
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/010.wav
- src: "ापखी खखैुेी"
- res: "ापखी खखैुेी"
--------------------------------------------------------------------------------
WER: 0.000000, CER: 0.000000, loss: 8.599924
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/008.wav
- src: "पेांीखतहाखाख खैुेी"
- res: "पेांीखतहाखाख खैुेी"
--------------------------------------------------------------------------------
WER: 0.000000, CER: 0.000000, loss: 11.154081
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/007.wav
- src: "ुबखंेदयदी"
- res: "ुबखंेदयदी"
--------------------------------------------------------------------------------
WER: 0.000000, CER: 0.000000, loss: 12.389311
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/001.wav
- src: " तपुबखैीबखुखतखप"
- res: " तपुबखैीबखुखतखप"
--------------------------------------------------------------------------------
WER: 0.000000, CER: 0.000000, loss: 12.756799
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/006.wav
- src: "बडैखखुखतखैयीखखपर"
- res: "बडैखखुखतखैयीखखपर"
--------------------------------------------------------------------------------
WER: 0.000000, CER: 0.000000, loss: 17.487480
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/003.wav
- src: "ाुखतरखदेतदीधीखुखाहखाख खैुखुपखखपन"
- res: "ाुखतरखदेतदीधीखुखाहखाख खैुखुपखखपन"
--------------------------------------------------------------------------------
WER: 0.000000, CER: 0.000000, loss: 18.130671
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/002.wav
- src: "तरखबडैीखैीखखापखी खखैुखाै खखपन"
- res: "तरखबडैीखैीखखापखी खखैुखाै खखपनlike this i am getting, it is because of low data? i think src should be displayed as it is.
I don’t understand, it looks like you have src == res, which would mean computed transcription matches expected transcription.
WER: 1.000000, CER: 0.600000, loss: 33.510384
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/009.wav
- src: "ैयीखखाखदुखखौखप हखपरा"
- res: "पखाख खपरा"
--------------------------------------------------------------------------------
WER: 0.000000, CER: 0.000000, loss: 6.128268
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/004.wav
- src: "ुखत"
- res: "ुखत"
--------------------------------------------------------------------------------
WER: 0.000000, CER: 0.000000, loss: 6.713559
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/005.wav
- src: "ापखी ख"
- res: "ापखी ख"
--------------------------------------------------------------------------------
WER: 0.000000, CER: 0.000000, loss: 7.111163
- wav: file:///home/yk/hindi-deep/DeepSpeech/data/test/010.wav
- src: "ापखी खखैुेी"
- res: "ापखी खखैुेी"
-------------------------------------------------------------------------
at first i got the WER 1
Well, this is the test set showing worst examples. I don’t see anything strange, please elaborate.
wav_filename,wav_filesize,transcript
001.wav,101000,तमहरआ कयआ नाम ह
002.wav,138000,मै आपकी कया सहायता कर सकता हू
003.wav,125000,सर मै विमवीशयोर से बात कर रहा हू
004.wav,78000,नाम
005.wav,99400,सहायता
006.wav,106000,आपका नाम क्या है
007.wav,80900,नई दिल्ली
008.wav,90700,हिंदी में बात करिए
009.wav,81900,क्या बोलना चाहते हैं
010.wav,88700,सहायता करिए
this is my original test.csv content match it with the src its totally different.
this is the original transcript
this is what i get in src