Could you upload your data somewhere? I could try to train a model and give you the config you need. Feels like that’s easier than debugging such specific issues.
@lucifera678
ok ok listen here. go to util/flags.py . change audio_sample_rate to 16000(you set it as 44100). and then you’ll see that you can export your model.
Do i know if that screws up your model? I do not. But can you export it? yes. good luck.
Is there restriction for wav file length for training?
You want to use data with max length 15 seconds, optimally even less than that (Mozilla model is trained on files with max length 8 seconds if I recall correctly).
More on how to deal with it here: Longer audio files with Deep Speech
Moreover in DeepSpeech source code in examples there are some scripts where vad transcriber was presented.
(venv) sehar@sehar-HP-Z220-CMT-Workstation:~/DeepSpeech$ python speech.py /home/sehar/urdu-models/output_graph1.pb /home/sehar/urdu-models/alphabet1.txt /home/sehar/urdu-models/sent6urd.wav
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.
2019-11-26 11:35:11.022262: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:998] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2019-11-26 11:35:11.023504: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties:
name: GeForce GTX 1060 6GB major: 6 minor: 1 memoryClockRate(GHz): 1.759
pciBusID: 0000:01:00.0
totalMemory: 5.93GiB freeMemory: 5.59GiB
2019-11-26 11:35:11.023541: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0
2019-11-26 11:35:15.973514: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:
2019-11-26 11:35:15.973566: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990] 0
2019-11-26 11:35:15.973593: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0: N
2019-11-26 11:35:16.001763: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 5371 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1060 6GB, pci bus id: 0000:01:00.0, compute capability: 6.1)
2019-11-26 11:35:17.972698: E tensorflow/core/framework/op_kernel.cc:1325] OpKernel (‘op: “UnwrapDatasetVariant” device_type: “CPU”’) for unknown op: UnwrapDatasetVariant
2019-11-26 11:35:17.972759: 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
2019-11-26 11:35:17.972786: E tensorflow/core/framework/op_kernel.cc:1325] OpKernel (‘op: “WrapDatasetVariant” device_type: “CPU”’) for unknown op: WrapDatasetVariant
2019-11-26 11:35:17.972954: 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
Error running session: Not found: PruneForTargets: Some target nodes not found: initialize_state
Segmentation fault (core dumped)
after training my model i tested it and its giving me this error
@sehar_capricon Please please please, can you really make an effort and USE CODE FORMATTING ? Your output is hard to read, this is DIFFICULT for people to help you.
You are running binary v0.5.1, your error would suggest you trained from current master which targets v0.6.x binaries.
thanks for your prompt response,
i have trained my model on deepspeech v 0.5.1
Can you triple check that ? How did you performed the export ?
you can improve the accuracy of your model by changing the audio file length to be between 15 to 20 seconds