Hindi accent using deepspeech

bharath.vadithya · November 18, 2019, 11:22am

after i run the script… i got something like this… i searched on internet, but iam unable to figure out it.

I STARTING Optimization
Epoch 0 |   Training | Elapsed Time: 0:00:00 | Steps: 0 | Loss: 0.000000                                                                                                                                   
Epoch 0 | Validation | Elapsed Time: 0:00:00 | Steps: 0 | Loss: 0.000000 | Dataset: minigir/train/train.csv                                                                                                
Traceback (most recent call last):
  File "DeepSpeech.py", line 931, in <module>
    absl.app.run(main)
  File "/home/metlife-vad/.local/lib/python3.7/site-packages/absl/app.py", line 299, in run
    _run_main(main, args)
  File "/home/metlife-vad/.local/lib/python3.7/site-packages/absl/app.py", line 250, in _run_main
    sys.exit(main(argv))
  File "DeepSpeech.py", line 915, in main
    train()
  File "DeepSpeech.py", line 642, in train
    dev_loss = dev_loss / total_steps
ZeroDivisionError: float division by zero

bharath.vadithya · November 18, 2019, 11:41am

hey @lissyx … i am just training on three examples will it enough for training (this is just to check weather iam getting results or not) so that i can go further training of more number data.

lissyx · November 18, 2019, 11:49am

What’s your training flags / command line ?

bharath.vadithya · November 18, 2019, 11:51am

i mean three audio files transcripts … the csv file has three audio transcripts

lissyx · November 18, 2019, 11:51am

Please, reply to what I asked.

bharath.vadithya · November 18, 2019, 11:52am

i didn’t understand !!

lissyx · November 18, 2019, 11:52am

Well then just say it. I need your python DeepSpeech.py [...] full command-line

bharath.vadithya · November 18, 2019, 11:53am

iam running this code

#!/usr/bin/env bash

set -xe
if [ ! -f DeepSpeech.py ]; then
    echo "Please make sure you run this from DeepSpeech's top level directory."
    exit 1
fi;

python3 -u DeepSpeech.py \
  --train_files minigir/train/train.csv \
  --dev_files minigir/train/train.csv \
  --test_files minigir/train/train.csv \
  --train_batch_size 48 \
  --dev_batch_size 40 \
  --test_batch_size 40 \
  --n_hidden 1024 \
  --epochs 64 \
  --early_stop True \
  --es_steps 6 \
  --es_mean_th 0.1 \
  --es_std_th 0.1 \
  --dropout_rate 0 \
  --log_level 1 \
  --learning_rate 0.000025 \
  --report_count 100 \
  --export_dir metlife-models/ \
  --checkpoint_dir metlife-models/check_point \
  --alphabet_config_path metlife-models/alphabet.txt \
  --lm_binary_path metlife-models/lm.binary \
  --lm_trie_path metlife-models/trie \
  "$@"

lissyx · November 18, 2019, 11:54am

Ok, if you only have three audio files, please try using batch size not above 3.

bharath.vadithya · November 18, 2019, 11:56am

okay … i got it …well i will try now

bharath.vadithya · November 18, 2019, 11:59am

it’s running … thank you @lissyx … btw how are you !

bharath.vadithya · November 25, 2019, 10:06am

hi @lissyx

Loading model from file metlife-models/output_graph.pb
TensorFlow: v1.13.1-10-g3e0cc53
DeepSpeech: v0.5.1-0-g4b29b78
Warning: reading entire model file into memory. Transform model file into an mmapped graph to reduce heap usage.
2019-11-25 15:24:22.279971: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
2019-11-25 15:24:22.320943: E tensorflow/core/framework/op_kernel.cc:1325] OpKernel ('op: "UnwrapDatasetVariant" device_type: "CPU"') for unknown op: UnwrapDatasetVariant
2019-11-25 15:24:22.321039: E tensorflow/core/framework/op_kernel.cc:1325] OpKernel ('op: "WrapDatasetVariant" device_type: "GPU" host_memory_arg: "input_handle" host_memory_arg: "output_handle"') for unknown op: WrapDatasetVariant
2019-11-25 15:24:22.321083: E tensorflow/core/framework/op_kernel.cc:1325] OpKernel ('op: "WrapDatasetVariant" device_type: "CPU"') for unknown op: WrapDatasetVariant
2019-11-25 15:24:22.321185: E tensorflow/core/framework/op_kernel.cc:1325] OpKernel ('op: "UnwrapDatasetVariant" device_type: "GPU" host_memory_arg: "input_handle" host_memory_arg: "output_handle"') for unknown op: UnwrapDatasetVariant
Specified model file version (0) is incompatible with minimum version supported by this client (1). See https://github.com/mozilla/DeepSpeech/#model-compatibility for more information
Traceback (most recent call last):
  File "/usr/local/bin/deepspeech", line 10, in <module>
    sys.exit(main())
  File "/usr/local/lib/python3.7/dist-packages/deepspeech/client.py", line 88, in main
    ds = Model(args.model, N_FEATURES, N_CONTEXT, args.alphabet, BEAM_WIDTH)
  File "/usr/local/lib/python3.7/dist-packages/deepspeech/__init__.py", line 23, in __init__
    raise RuntimeError("CreateModel failed with error code {}".format(status))
RuntimeError: CreateModel failed with error code 8195

i searched whole internet i didn’t find the solution for this and iam using correct deepspeech version

lissyx · November 25, 2019, 10:07am

You have the error here … You need to share more details, but it looks like you exported wrongly.

bharath.vadithya · November 25, 2019, 10:11am

sudo deepspeech --model metlife-models/output_graph.pb --alphabet metlife-models/alphabet.txt --lm metlife-models/lm.binary --trie metlife-models/trie --audio minigir/wav/tmp2.wav 
Loading model from file metlife-models/output_graph.pb
TensorFlow: v1.13.1-10-g3e0cc53
DeepSpeech: v0.5.1-0-g4b29b78
Warning: reading entire model file into memory. Transform model file into an mmapped graph to reduce heap usage.
2019-11-25 15:24:22.279971: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
2019-11-25 15:24:22.320943: E tensorflow/core/framework/op_kernel.cc:1325] OpKernel ('op: "UnwrapDatasetVariant" device_type: "CPU"') for unknown op: UnwrapDatasetVariant
2019-11-25 15:24:22.321039: E tensorflow/core/framework/op_kernel.cc:1325] OpKernel ('op: "WrapDatasetVariant" device_type: "GPU" host_memory_arg: "input_handle" host_memory_arg: "output_handle"') for unknown op: WrapDatasetVariant
2019-11-25 15:24:22.321083: E tensorflow/core/framework/op_kernel.cc:1325] OpKernel ('op: "WrapDatasetVariant" device_type: "CPU"') for unknown op: WrapDatasetVariant
2019-11-25 15:24:22.321185: E tensorflow/core/framework/op_kernel.cc:1325] OpKernel ('op: "UnwrapDatasetVariant" device_type: "GPU" host_memory_arg: "input_handle" host_memory_arg: "output_handle"') for unknown op: UnwrapDatasetVariant
Specified model file version (0) is incompatible with minimum version supported by this client (1). See https://github.com/mozilla/DeepSpeech/#model-compatibility for more information
Traceback (most recent call last):
  File "/usr/local/bin/deepspeech", line 10, in <module>
    sys.exit(main())
  File "/usr/local/lib/python3.7/dist-packages/deepspeech/client.py", line 88, in main
    ds = Model(args.model, N_FEATURES, N_CONTEXT, args.alphabet, BEAM_WIDTH)
  File "/usr/local/lib/python3.7/dist-packages/deepspeech/__init__.py", line 23, in __init__
    raise RuntimeError("CreateModel failed with error code {}".format(status))
RuntimeError: CreateModel failed with error code 8195

what is model compatability and error code 8195. still did’t understand

lissyx · November 25, 2019, 10:12am

Ok, seriously, read the links, share the informations I am asking.

Please avoid using sudo when it’s not necessary.

bharath.vadithya · November 25, 2019, 10:13am

iam using sudo for some permission to access

lissyx · November 25, 2019, 10:16am

Your setup is likely wrong, there’s absolutely no reason you should have to do this …

lissyx · November 25, 2019, 10:18am

Model version is documented and checking is here to ensure you don’t try to run a model that is not compatible with a binary. SInce you have still not documented your export phase, I cannot help you. And since we are now close to 100 messages to help you I’m really getting close to the end of my patience.

github.com

mozilla/DeepSpeech/blob/master/USING.rst#model-compatibility

Using a Pre-trained Model
=========================

Inference using a DeepSpeech pre-trained model can be done with a client/language binding package. We have four clients/language bindings in this repository, listed below, and also a few community-maintained clients/language bindings in other repositories, listed `further down in this README <#third-party-bindings>`_.


* `The Python package/language binding <#using-the-python-package>`_
* `The Node.JS package/language binding <#using-the-nodejs-package>`_
* `The Command-Line client <#using-the-command-line-client>`_
* `The .NET client/language binding <native_client/dotnet/README.rst>`_

Running ``deepspeech`` might, see below, require some runtime dependencies to be already installed on your system:


* sox - The Python and Node.JS clients use SoX to resample files to 16kHz.
* libgomp1 - libsox (statically linked into the clients) depends on OpenMP. Some people have had to install this manually.
* libstdc++ - Standard C++ Library implementation. Some people have had to install this manually.
* libpthread - On Linux, some people have had to install libpthread manually.

Please refer to your system's documentation on how to install these dependencies.

This file has been truncated. show original

https://deepspeech.readthedocs.io/en/latest/Error-Codes.html

So please make an effort.

bharath.vadithya · November 25, 2019, 10:19am

i tried to upgrade to newest model of deepspeech it still showing the same error.

how can i downgrade or re-export it

lissyx · November 25, 2019, 10:19am

Which is both expected and not useful, since you don’t document “newest model”, neither “upgrade to”, and I still don’t know how you performed your failing export, nor your current setup.