What and how to report if you need support

We are happy for every new user of DeepSpeech. As it is under active development, stuff changes quickly and you might have problems running it. We are here to help and might point to this thread if needed :slight_smile:

Please do the following before posting:

If you still struggle, feel free to open a new thread in this forum (don’t answer to this post). Only hijack older threads if really necessary.

Please include the following when reporting a problem:

  • Training or just running inference
  • Moziall STT branch/version
  • OS Platform and Distribution (e.g., Linux Ubuntu 18.04)
  • Python version
  • TensorFlow version

Why not try GatherUp to collect and pretty print most of that?

If using GPU:

  • CUDA/cuDNN version

State what you have learned from reading older threads. Usually there is already some discussion about the problem. You might have a new angle and it is great to begin talking about it with you having some background of what was said before.

Include logs (as citation/preformatted) so we can see the formatted output. No images please.

We love to hear what you plan on doing with DeepSpeech. Usually you’ll get additional feedback or hints on how to succeed.

Finally, sometimes we can respond within seconds, sometimes latency can be measured in days. Remember this is free

Is there a deepspeech v0.7.4 pretrained model for German and Mandarin?
Question about scorer
Generate_lm.py error when reproducing external scorer
Plot a training loss graph
[Python] ValueError: invalid literal for int() with base 10
Training on Urdu giving eager execution error
Creating tflite file for Android
DeepSpeech model training accuracy
Deepspeech not recognizing Indian English Accent
Using common voice datasets?
Zero nodes when converting to memmapped format. Deepspeech
Add new language
Generating own scorer file
How to create a mmap-able model from the output_graph.pb file
Mprove speech to text deep speech
Problem when training my own model
Missing character
Is the language model based on character or word?
Need help installing DeepSpeech for training/transfer learning in Google Cloud VM
Model is inferencing bad after finetuning
Alphabet cannot encode transcript
DeepSpeech 0.8.1 Blank inference returned
Do DeepSpeech have subtitle (SRT) output mode? How can I merge words into the proper sentences?
How to solve stackoverflow exception in .NET library
Generate_lm error
Division by zero in deep speech
TUTORIAL : How I trained a specific french model to control my robot
How to train Deepspeech model on google colab?
No --validate_label_locale specified, your might end with inconsistent dataset
Install failed, using commands from DP Documentation
Fine Tuning with limited data - Questions on Fine Tuning in General
Error when trying to decode: terminate called after throwing an instance of 'std::length_error'
Tune MoziilaDeepSpeech to recognize specific sentences
Can someone explain how to prepare my own data (format of both audio and transcription)?
Check result of some Augmentation techniques
TUTORIAL : How I trained a specific french model to control my robot
Running taskcluster.py : “IO error [errno socket error] Name or service not known)
Problem in Convergence - DeepSpeech not learning
Cuda is not getting load for DeepSpeech-GPU 0.7.4
Transcribe.py wrong transcription
Invalid String
Deepspeech 0.7.4 installation settings
Generate_lm.py subprocess.CalledProcessError
Optimaizatiom the WER result
Fine-tuning in chinese model and loss suddenly increases
lm::FormatLoadException Error on SageMaker with Persistent Environment
Problem when transcript from microphone
I am getting Validation Loss: inf