Open Koowater opened 11 months ago
rybakov
commented
11 months ago I fixed bc_resnet with 'causal' padding and added a test with 'bc_resnet_causal'.
You could apply Delay layer to bc_resnet the same way it is done in delay_test.py has several examples combining conv with delay layers.
It can be easier to re-design it using sub class api as shown in example
Koowater
commented
11 months ago Thank you for answering my issue, @rybakov.
I'm trying to apply causal padding and Delay layer for residual and identity connection. But I got an error when I convert my model to tflite_streaming_model.
I'll try to find out where I'm going wrong, but I don't know how to fix it at the moment. Could you please review my errors?
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 2
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 2
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 4
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 4
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 4
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 8
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 8
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 8
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 8
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 8
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 16
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 16
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 16
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 16
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 16
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
input_1 (InputLayer) [(1, 16000)] 0
speech_features (SpeechFeat (1, 98, 40) 0
ures)
tf_op_layer_ExpandDims (Ten [(1, 98, 40, 1)] 0
sorFlowOpLayer)
stream (Stream) (1, 98, 20, 16) 416
transition_block (Transitio (1, 98, 20, 8) 464
nBlock)
normal_block (NormalBlock) (1, 98, 20, 8) 304
normal_block_1 (NormalBlock (1, 98, 20, 8) 304
)
transition_block_1 (Transit (1, 98, 10, 12) 648
ionBlock)
normal_block_2 (NormalBlock (1, 98, 10, 12) 504
)
normal_block_3 (NormalBlock (1, 98, 10, 12) 504
)
transition_block_2 (Transit (1, 98, 5, 16) 992
ionBlock)
normal_block_4 (NormalBlock (1, 98, 5, 16) 736
)
normal_block_5 (NormalBlock (1, 98, 5, 16) 736
)
normal_block_6 (NormalBlock (1, 98, 5, 16) 736
)
normal_block_7 (NormalBlock (1, 98, 5, 16) 736
)
transition_block_3 (Transit (1, 98, 5, 20) 1400
ionBlock)
normal_block_8 (NormalBlock (1, 98, 5, 20) 1000
)
normal_block_9 (NormalBlock (1, 98, 5, 20) 1000
)
normal_block_10 (NormalBloc (1, 98, 5, 20) 1000
k)
normal_block_11 (NormalBloc (1, 98, 5, 20) 1000
k)
stream_33 (Stream) (1, 98, 5, 20) 520
tf_op_layer_Mean (TensorFlo [(1, 98, 1, 20)] 0
wOpLayer)
conv2d_21 (Conv2D) (1, 98, 1, 32) 640
stream_34 (Stream) (1, 1, 1, 32) 0
conv2d_22 (Conv2D) (1, 1, 1, 12) 384
tf_op_layer_Squeeze (Tensor [(1, 12)] 0
FlowOpLayer)
=================================================================
Total params: 14,024
Trainable params: 11,032
Non-trainable params: 2,992
_________________________________________________________________
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 5
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 2
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 2
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 4
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 4
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 8
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 8
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 8
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 8
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 16
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 16
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 16
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 16
WARNING:absl:ring_buffer_size_in_time_dim overwritten by the passed-in value: 5
Traceback (most recent call last):
File "decode.py", line 196, in <module>
tflite_streaming_model = utils.model_to_tflite(sess, model_non_stream_batch, flags, Modes.STREAM_EXTERNAL_STATE_INFERENCE)
File "/tf/kws/koowater/builder/kws_streaming/models/utils.py", line 386, in model_to_tflite
model_stream = to_streaming_inference(model_non_stream, flags, mode)
File "/tf/kws/koowater/builder/kws_streaming/models/utils.py", line 318, in to_streaming_inference
model_inference = convert_to_inference_model(model_non_stream,
File "/tf/kws/koowater/builder/kws_streaming/models/utils.py", line 249, in convert_to_inference_model
new_model = _clone_model(model, input_tensors)
File "/tf/kws/koowater/builder/kws_streaming/models/utils.py", line 109, in _clone_model
functional.reconstruct_from_config(
File "/usr/local/lib/python3.8/dist-packages/keras/engine/functional.py", line 1495, in reconstruct_from_config
if process_node(layer, node_data):
File "/usr/local/lib/python3.8/dist-packages/keras/engine/functional.py", line 1435, in process_node
output_tensors = layer(input_tensors, **kwargs)
File "/usr/local/lib/python3.8/dist-packages/keras/engine/base_layer_v1.py", line 838, in __call__
outputs = call_fn(cast_inputs, *args, **kwargs)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/impl/api.py", line 692, in wrapper
raise e.ag_error_metadata.to_exception(e)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/impl/api.py", line 689, in wrapper
return converted_call(f, args, kwargs, options=options)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/impl/api.py", line 439, in converted_call
result = converted_f(*effective_args, **kwargs)
File "/tmp/__autograph_generated_fileizg6by4c.py", line 105, in tf__call
ag__.if_stmt((ag__.ld(self).mode == ag__.ld(modes).Modes.STREAM_INTERNAL_STATE_INFERENCE), if_body_4, else_body_4, get_state_4, set_state_4, ('do_return', 'retval_', 'self.output_state'), 3)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/operators/control_flow.py", line 1266, in if_stmt
_py_if_stmt(cond, body, orelse)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/operators/control_flow.py", line 1319, in _py_if_stmt
return body() if cond else orelse()
File "/tmp/__autograph_generated_fileizg6by4c.py", line 103, in else_body_4
ag__.if_stmt((ag__.ld(self).mode == ag__.ld(modes).Modes.STREAM_EXTERNAL_STATE_INFERENCE), if_body_3, else_body_3, get_state_3, set_state_3, ('do_return', 'retval_', 'self.output_state'), 3)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/operators/control_flow.py", line 1266, in if_stmt
_py_if_stmt(cond, body, orelse)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/operators/control_flow.py", line 1319, in _py_if_stmt
return body() if cond else orelse()
File "/tmp/__autograph_generated_fileizg6by4c.py", line 71, in if_body_3
ag__.if_stmt(ag__.ld(self).ring_buffer_size_in_time_dim, if_body_1, else_body_1, get_state_1, set_state_1, ('output', 'self.output_state'), 2)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/operators/control_flow.py", line 1266, in if_stmt
_py_if_stmt(cond, body, orelse)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/operators/control_flow.py", line 1319, in _py_if_stmt
return body() if cond else orelse()
File "/tmp/__autograph_generated_fileizg6by4c.py", line 65, in if_body_1
(output, ag__.ld(self).output_state) = ag__.converted_call(ag__.ld(self)._streaming_external_state, (ag__.ld(inputs), ag__.ld(self).input_state), None, fscope)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/impl/api.py", line 441, in converted_call
result = converted_f(*effective_args)
File "/tmp/__autograph_generated_file_lmun91g.py", line 166, in tf___streaming_external_state
ag__.if_stmt(ag__.converted_call(ag__.ld(isinstance), (ag__.converted_call(ag__.ld(self).get_core_layer, (), None, fscope), ag__.ld(tf).keras.layers.Conv2DTranspose), None, fscope), if_body_6, else_body_6, get_state_6, set_state_6, ('do_return', 'retval_'), 2)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/operators/control_flow.py", line 1266, in if_stmt
_py_if_stmt(cond, body, orelse)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/operators/control_flow.py", line 1319, in _py_if_stmt
return body() if cond else orelse()
File "/tmp/__autograph_generated_file_lmun91g.py", line 157, in else_body_6
ag__.if_stmt(ag__.ld(self).use_one_step, if_body_5, else_body_5, get_state_5, set_state_5, ('do_return', 'retval_'), 2)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/operators/control_flow.py", line 1266, in if_stmt
_py_if_stmt(cond, body, orelse)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/operators/control_flow.py", line 1319, in _py_if_stmt
return body() if cond else orelse()
File "/tmp/__autograph_generated_file_lmun91g.py", line 134, in if_body_5
memory = ag__.converted_call(ag__.ld(tf).keras.backend.concatenate, ([ag__.ld(memory), ag__.ld(inputs)], 1), None, fscope)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/impl/api.py", line 331, in converted_call
return _call_unconverted(f, args, kwargs, options, False)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/autograph/impl/api.py", line 459, in _call_unconverted
return f(*args)
File "/usr/local/lib/python3.8/dist-packages/tensorflow/python/util/traceback_utils.py", line 153, in error_handler
raise e.with_traceback(filtered_tb) from None
File "/usr/local/lib/python3.8/dist-packages/keras/backend.py", line 3581, in concatenate
return tf.concat([to_dense(x) for x in tensors], axis)
ValueError: in user code:
File "/tf/kws/koowater/builder/kws_streaming/layers/stream.py", line 411, in call *
output, self.output_state = self._streaming_external_state(
File "/tf/kws/koowater/builder/kws_streaming/layers/stream.py", line 563, in _streaming_external_state *
memory = tf.keras.backend.concatenate([memory, inputs], 1)
File "/usr/local/lib/python3.8/dist-packages/keras/backend.py", line 3581, in concatenate
return tf.concat([to_dense(x) for x in tensors], axis)
ValueError: Dimension 0 in both shapes must be equal, but are 40 and 43. Shapes are [40,1] and [43,1]. for '{{node streaming/stream/concat}} = ConcatV2[N=2, T=DT_FLOAT, Tidx=DT_INT32](streaming/stream/strided_slice, streaming/stream/Pad, streaming/stream/concat/axis)' with input shapes: [1,4,40,1], [1,1,43,1], [] and with computed input tensors: input[2] = <1>.I used Modes.STREAM_EXTERNAL_STATE_INFERENCE and I found that my error is raised in first Conv2D layer after expand_dim. frequency_pad in Stream layer occur frequency dimension mismatch. In training, this frequency dimension was applied mean layer, so this is not important, but in STREAM_EXTERNAL_STATE_INFERENCE this mismatch occur error. This issue may be my fault, please review this issue.
[+] I changed inference mode to STREAM_INTERNAL_STATE_INFERENCE and conversion is works. But I could not believe my model works well because streaming model's output and non streaming model's output is different...
rybakov
commented
11 months ago Please confirm that you pulled the latest version of kws_streaming
Koowater
commented
11 months ago @rybakov Sorry, and thank you so much! I pulled the latest version of kws_streaming and bc_resnet was converted to streaming inference well.
But I have a last question, model's prediction is difference between streaming inference and non streaming inference. I want to know bc_resnet's input data_shape at streaming inference. I used data_shape = (160,).
Here's my output.
*** Here is streaming inference ***
[silence]
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
[silence]
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
[silence]
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
[silence]
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
[silence]
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
{1: '_unknown_', 2: 'yes', 4: 'up', 10: 'stop', 3: 'no', 11: 'go', 5: 'down', 9: 'off', 8: 'on', 6: 'left', 7: 'right', 0: '_silence_'}
../datasets/data2/yes/1b88bf70_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/yes/05b2db80_nohash_1.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 0, 0, 0, 0]
../datasets/data2/yes/b66f4f93_nohash_5.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6]
../datasets/data2/yes/750e3e75_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 11, 11, 11, 11, 11, 11, 11, 11, 11, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/yes/e49428d9_nohash_3.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/yes/af7a8296_nohash_3.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9]
../datasets/data2/yes/778a4a01_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 6, 6, 6, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/yes/b00dff7e_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6]
../datasets/data2/yes/e77d88fc_nohash_1.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[6, 6, 6, 6, 6, 6, 6, 6, 6, 9, 9, 9, 9, 9, 9, 9, 9, 9, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10]
../datasets/data2/yes/0cb74144_nohash_2.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 5, 5, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/yes/3d794813_nohash_4.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/yes/11321027_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6]
../datasets/data2/yes/06f6c194_nohash_4.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 11, 11, 11, 11, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]
../datasets/data2/yes/b97c9f77_nohash_1.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/yes/7213ed54_nohash_4.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/yes/c50f55b8_nohash_5.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 11, 11, 11, 11, 11, 11, 2, 2, 2, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/yes/09bcdc9d_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 6, 6, 6, 6, 6, 6, 6, 6, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]
../datasets/data2/yes/cae62f38_nohash_1.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]
../datasets/data2/yes/1942abd7_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 11, 11, 11, 11, 11, 11, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]
../datasets/data2/yes/321aba74_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 11, 11, 11, 11, 11, 11, 0, 0, 0, 0, 0, 0, 0, 11, 11, 11, 11, 2, 2, 2, 2, 2, 2, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/no/b66f4f93_nohash_5.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6]
../datasets/data2/no/66cbe2b3_nohash_2.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 9, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 4, 4, 4, 4, 4, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/no/750e3e75_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 4, 4, 4, 4, 4, 4, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 11, 11, 4, 4, 4, 4, 11, 11, 11, 11, 11, 11, 11, 11, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6]
../datasets/data2/no/3852fca2_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 11, 11, 11, 11, 11, 11, 11, 11, 6, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/no/e49428d9_nohash_3.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 4, 4, 4, 4, 4, 4, 4, 4, 11, 11, 11, 11, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4]
../datasets/data2/no/778a4a01_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/no/b00dff7e_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 6, 6, 6, 6, 6, 6, 4, 4, 4, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4]
../datasets/data2/no/61e50f62_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4]
../datasets/data2/no/e77d88fc_nohash_1.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 9, 9, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11]
../datasets/data2/no/0cb74144_nohash_2.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 11, 11, 11, 11, 11, 11, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4]
../datasets/data2/no/17c94b23_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4]
../datasets/data2/no/3d794813_nohash_4.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9]
../datasets/data2/no/e55a2b20_nohash_1.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9]
../datasets/data2/no/06f6c194_nohash_4.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9]
../datasets/data2/no/b97c9f77_nohash_1.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6]
../datasets/data2/no/7213ed54_nohash_4.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6]
../datasets/data2/no/c50f55b8_nohash_5.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6]
../datasets/data2/no/09bcdc9d_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 4, 4, 4, 4, 4, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 9, 9, 9, 9, 9, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4]
../datasets/data2/no/1942abd7_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4]
../datasets/data2/no/321aba74_nohash_0.wav
[input_data.shape] (1, 16000) [out_tflite.shape] (100, 12)
[4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 11, 11, 11, 11, 11, 11, 11, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6]
*** Here is non streaming inference ***
yes
../datasets/data2/yes/05b2db80_nohash_1.wav
yes
../datasets/data2/yes/b66f4f93_nohash_5.wav
yes
../datasets/data2/yes/750e3e75_nohash_0.wav
yes
../datasets/data2/yes/e49428d9_nohash_3.wav
yes
../datasets/data2/yes/af7a8296_nohash_3.wav
yes
../datasets/data2/yes/778a4a01_nohash_0.wav
yes
../datasets/data2/yes/b00dff7e_nohash_0.wav
yes
../datasets/data2/yes/e77d88fc_nohash_1.wav
yes
../datasets/data2/yes/0cb74144_nohash_2.wav
yes
../datasets/data2/yes/3d794813_nohash_4.wav
yes
../datasets/data2/yes/11321027_nohash_0.wav
yes
../datasets/data2/yes/06f6c194_nohash_4.wav
yes
../datasets/data2/yes/b97c9f77_nohash_1.wav
yes
../datasets/data2/yes/7213ed54_nohash_4.wav
yes
../datasets/data2/yes/c50f55b8_nohash_5.wav
yes
../datasets/data2/yes/09bcdc9d_nohash_0.wav
yes
../datasets/data2/yes/cae62f38_nohash_1.wav
yes
../datasets/data2/yes/1942abd7_nohash_0.wav
yes
../datasets/data2/yes/321aba74_nohash_0.wav
yes
../datasets/data2/no/b66f4f93_nohash_5.wav
no
../datasets/data2/no/66cbe2b3_nohash_2.wav
no
../datasets/data2/no/750e3e75_nohash_0.wav
no
../datasets/data2/no/3852fca2_nohash_0.wav
no
../datasets/data2/no/e49428d9_nohash_3.wav
no
../datasets/data2/no/778a4a01_nohash_0.wav
no
../datasets/data2/no/b00dff7e_nohash_0.wav
no
../datasets/data2/no/61e50f62_nohash_0.wav
no
../datasets/data2/no/e77d88fc_nohash_1.wav
no
../datasets/data2/no/0cb74144_nohash_2.wav
no
../datasets/data2/no/17c94b23_nohash_0.wav
no
../datasets/data2/no/3d794813_nohash_4.wav
no
../datasets/data2/no/e55a2b20_nohash_1.wav
no
../datasets/data2/no/06f6c194_nohash_4.wav
no
../datasets/data2/no/b97c9f77_nohash_1.wav
no
../datasets/data2/no/7213ed54_nohash_4.wav
no
../datasets/data2/no/c50f55b8_nohash_5.wav
no
../datasets/data2/no/09bcdc9d_nohash_0.wav
no
../datasets/data2/no/1942abd7_nohash_0.wav
no
../datasets/data2/no/321aba74_nohash_0.wav
no
krilina988
commented
4 months ago Hello, how did you modify bc_resnet to enable streaming inference? I refer to delay_test and the modification always fails. I don't know what went wrong. @Koowater @rybakov
I'm trying to apply Delay layer in bc_resnet for streaming in 'same' padding. Because I got an error when I used 'causal' padding to bc_resnet...
I'm wondering how to apply Delay layer to bc_resnet. Delay layer needs a delay_val, and delay_val is calculated using kernel_size and dilation. But bc_resnet's kernel_size and dilation is not integer, these are tuple.
How to apply Delay layer in bc_resnet?
(Sorry, I'm seeing #1425 but I need more definite answer...)