Created
December 26, 2019 06:47
-
-
Save manish-kumar-garg/ab616732fb286c61162b03db9732358a to your computer and use it in GitHub Desktop.
Returnn - asr_local_attention
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
RETURNN starting up, version 20191217.234858--git-09b41c6f-dirty, date/time 2019-12-26-06-45-58 (UTC+0000), pid 20215, cwd /home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention, Python /home/ubuntu/tf1.13/bin/python3 | |
RETURNN command line options: ['local-heuristic.argmax.win05.exp3.ctc.config'] | |
Hostname: ip-10-1-21-241 | |
/home/ubuntu/tf1.13/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:526: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'. | |
_np_qint8 = np.dtype([("qint8", np.int8, 1)]) | |
/home/ubuntu/tf1.13/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:527: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'. | |
_np_quint8 = np.dtype([("quint8", np.uint8, 1)]) | |
/home/ubuntu/tf1.13/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:528: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'. | |
_np_qint16 = np.dtype([("qint16", np.int16, 1)]) | |
/home/ubuntu/tf1.13/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:529: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'. | |
_np_quint16 = np.dtype([("quint16", np.uint16, 1)]) | |
/home/ubuntu/tf1.13/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:530: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'. | |
_np_qint32 = np.dtype([("qint32", np.int32, 1)]) | |
/home/ubuntu/tf1.13/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:535: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'. | |
np_resource = np.dtype([("resource", np.ubyte, 1)]) | |
TensorFlow: 1.13.1 (b'v1.13.1-0-g6612da8951') (<site-package> in /home/ubuntu/tf1.13/lib/python3.6/site-packages/tensorflow) | |
Setup TF inter and intra global thread pools, num_threads None, session opts {'log_device_placement': False, 'device_count': {'GPU': 0}}. | |
2019-12-26 06:45:59.202047: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA | |
2019-12-26 06:45:59.425681: 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 | |
2019-12-26 06:45:59.429258: 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 | |
2019-12-26 06:45:59.430831: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x55e4d8dd5850 executing computations on platform CUDA. Devices: | |
2019-12-26 06:45:59.430864: I tensorflow/compiler/xla/service/service.cc:158] StreamExecutor device (0): Tesla K80, Compute Capability 3.7 | |
2019-12-26 06:45:59.430881: I tensorflow/compiler/xla/service/service.cc:158] StreamExecutor device (1): Tesla K80, Compute Capability 3.7 | |
2019-12-26 06:45:59.450542: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2300090000 Hz | |
2019-12-26 06:45:59.452625: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x55e4d9498d60 executing computations on platform Host. Devices: | |
2019-12-26 06:45:59.452659: I tensorflow/compiler/xla/service/service.cc:158] StreamExecutor device (0): <undefined>, <undefined> | |
2019-12-26 06:45:59.452768: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix: | |
2019-12-26 06:45:59.452791: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990] | |
CUDA_VISIBLE_DEVICES is set to '1,2'. | |
Collecting TensorFlow device list... | |
2019-12-26 06:45:59.455892: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties: | |
name: Tesla K80 major: 3 minor: 7 memoryClockRate(GHz): 0.8235 | |
pciBusID: 0000:00:18.0 | |
totalMemory: 11.17GiB freeMemory: 446.06MiB | |
2019-12-26 06:45:59.456040: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 1 with properties: | |
name: Tesla K80 major: 3 minor: 7 memoryClockRate(GHz): 0.8235 | |
pciBusID: 0000:00:19.0 | |
totalMemory: 11.17GiB freeMemory: 446.06MiB | |
2019-12-26 06:45:59.456482: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0, 1 | |
2019-12-26 06:45:59.459962: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix: | |
2019-12-26 06:45:59.459989: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990] 0 1 | |
2019-12-26 06:45:59.460007: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0: N Y | |
2019-12-26 06:45:59.460016: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 1: Y N | |
2019-12-26 06:45:59.460292: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/device:GPU:0 with 221 MB memory) -> physical GPU (device: 0, name: Tesla K80, pci bus id: 0000:00:18.0, compute capability: 3.7) | |
2019-12-26 06:45:59.463035: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/device:GPU:1 with 221 MB memory) -> physical GPU (device: 1, name: Tesla K80, pci bus id: 0000:00:19.0, compute capability: 3.7) | |
Local devices available to TensorFlow: | |
1/6: name: "/device:CPU:0" | |
device_type: "CPU" | |
memory_limit: 268435456 | |
locality { | |
} | |
incarnation: 13382603103062640486 | |
2/6: name: "/device:XLA_GPU:0" | |
device_type: "XLA_GPU" | |
memory_limit: 17179869184 | |
locality { | |
} | |
incarnation: 13903808377023195374 | |
physical_device_desc: "device: XLA_GPU device" | |
3/6: name: "/device:XLA_GPU:1" | |
device_type: "XLA_GPU" | |
memory_limit: 17179869184 | |
locality { | |
} | |
incarnation: 10467231846590301634 | |
physical_device_desc: "device: XLA_GPU device" | |
4/6: name: "/device:XLA_CPU:0" | |
device_type: "XLA_CPU" | |
memory_limit: 17179869184 | |
locality { | |
} | |
incarnation: 2300646191363693310 | |
physical_device_desc: "device: XLA_CPU device" | |
5/6: name: "/device:GPU:0" | |
device_type: "GPU" | |
memory_limit: 231800832 | |
locality { | |
bus_id: 1 | |
links { | |
link { | |
device_id: 1 | |
type: "StreamExecutor" | |
strength: 1 | |
} | |
} | |
} | |
incarnation: 15532672733655798053 | |
physical_device_desc: "device: 0, name: Tesla K80, pci bus id: 0000:00:18.0, compute capability: 3.7" | |
6/6: name: "/device:GPU:1" | |
device_type: "GPU" | |
memory_limit: 231800832 | |
locality { | |
bus_id: 1 | |
links { | |
link { | |
type: "StreamExecutor" | |
strength: 1 | |
} | |
} | |
} | |
incarnation: 4180441292150800506 | |
physical_device_desc: "device: 1, name: Tesla K80, pci bus id: 0000:00:19.0, compute capability: 3.7" | |
Using gpu device 1: Tesla K80 | |
Using gpu device 2: Tesla K80 | |
<LibriSpeechCorpus 'train' epoch=1>, epoch 1. Old mean seq len (transcription) is 183.267376, new is 63.708029, requested max is 75.000000. Old num seqs is 6575, new num seqs is 822. | |
<LibriSpeechCorpus 'train' epoch=1>, epoch 1. Old num seqs 14063, new num seqs 822. | |
<LibriSpeechCorpus 'train' epoch=1>, epoch 1. Old mean seq len (transcription) is 183.267376, new is 63.708029, requested max is 75.000000. Old num seqs is 6575, new num seqs is 822. | |
<LibriSpeechCorpus 'train' epoch=1>, epoch 1. Old num seqs 14063, new num seqs 822. | |
Train data: | |
input: 40 x 1 | |
output: {'classes': [10025, 1], 'raw': {'dtype': 'string', 'shape': ()}, 'data': [40, 2]} | |
LibriSpeechCorpus, sequences: 822, frames: unknown | |
Dev data: | |
LibriSpeechCorpus, sequences: 3000, frames: unknown | |
Learning-rate-control: file newbob.data does not exist yet | |
Update config key 'max_seq_length' for epoch 1: {'classes': 75} -> {'classes': 60} | |
Setup tf.Session with options {'log_device_placement': False, 'device_count': {'GPU': 1}} ... | |
2019-12-26 06:46:05.591177: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0, 1 | |
2019-12-26 06:46:05.591272: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix: | |
2019-12-26 06:46:05.591293: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990] 0 1 | |
2019-12-26 06:46:05.591310: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0: N Y | |
2019-12-26 06:46:05.591325: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 1: Y N | |
2019-12-26 06:46:05.591489: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 221 MB memory) -> physical GPU (device: 0, name: Tesla K80, pci bus id: 0000:00:18.0, compute capability: 3.7) | |
layer root/'data' output: Data(name='data', shape=(None, 40), batch_shape_meta=[B,T|'time:var:extern_data:data',F|40]) | |
layer root/'source' output: Data(name='source_output', shape=(None, 40), batch_shape_meta=[B,T|'time:var:extern_data:data',F|40]) | |
layer root/'lstm0_fw' output: Data(name='lstm0_fw_output', shape=(None, 1024), batch_dim_axis=1, batch_shape_meta=[T|'time:var:extern_data:data',B,F|1024]) | |
layer root/'lstm0_bw' output: Data(name='lstm0_bw_output', shape=(None, 1024), batch_dim_axis=1, batch_shape_meta=[T|'time:var:extern_data:data',B,F|1024]) | |
layer root/'lstm0_pool' output: Data(name='lstm0_pool_output', shape=(None, 2048), batch_shape_meta=[B,T|?,F|2048]) | |
layer root/'lstm5_fw' output: Data(name='lstm5_fw_output', shape=(None, 1024), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|1024]) | |
layer root/'lstm5_bw' output: Data(name='lstm5_bw_output', shape=(None, 1024), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|1024]) | |
layer root/'encoder' output: Data(name='encoder_output', shape=(None, 2048), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|2048]) | |
layer root/'ctc' output: Data(name='ctc_output', shape=(None, 10026), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|10026]) | |
layer root/'enc_ctx' output: Data(name='enc_ctx_output', shape=(None, 1024), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|1024]) | |
layer root/'inv_fertility' output: Data(name='inv_fertility_output', shape=(None, 1), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|1]) | |
layer root/'enc_value' output: Data(name='enc_value_output', shape=(None, 1, 2048), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,1,F|2048]) | |
<_SubnetworkRecCell of None>: exception constructing template network (for deps and data shapes) | |
Most recent construction stack: | |
<_TemplateLayer(EvalLayer)(:template:eval) 'output/p_t_in' out_type=Data(shape=(), time_dim_axis=None, batch_shape_meta=[B]) (construction stack 'att_weights')>, kwargs: | |
{'eval': 'tf.squeeze(tf.argmax(source(0), axis=1, output_type=tf.int32), ' | |
'axis=1)', | |
'name': 'p_t_in', | |
'network': <TFNetwork 'root/output:rec-subnet' parent_net=<TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>>, | |
'out_type': {'batch_dim_axis': 0, 'dtype': 'float32', 'shape': ()}, | |
'sources': [<_TemplateLayer(SoftmaxOverSpatialLayer)(:prev:softmax_over_spatial) 'output/prev:att_weights' out_type=Data(shape=(1, None), time_dim_axis=2, feature_dim_axis=1, batch_shape_meta=[B,F|1,T|'spatial:0:lstm0_pool']) (construction stack None)>]} | |
Template network so far: | |
{'accum_att_weights': <_TemplateLayer(EvalLayer)(:template:eval) 'output/accum_att_weights' out_type=Data(shape=(None, 1), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|1]) (construction stack 'weight_feedback')>, | |
'att': <_TemplateLayer(MergeDimsLayer)(:template:merge_dims) 'output/att' out_type=Data(shape=(2048,), time_dim_axis=None, batch_shape_meta=[B,F|2048]) (construction stack 's')>, | |
'att0': <_TemplateLayer(GenericAttentionLayer)(:template:generic_attention) 'output/att0' out_type=Data(shape=(1, 2048), time_dim_axis=None, batch_shape_meta=[B,1,F|2048]) (construction stack 'att')>, | |
'att_weights': <_TemplateLayer(SoftmaxOverSpatialLayer)(:template:softmax_over_spatial) 'output/att_weights' out_type=Data(shape=(1, None), time_dim_axis=2, feature_dim_axis=1, batch_shape_meta=[B,F|1,T|'spatial:0:lstm0_pool']) (construction stack 'att0')>, | |
'end': <_TemplateLayer(CompareLayer)(:template:compare) 'output/end' out_type=Data(shape=(), dtype='bool', sparse=True, dim=2, time_dim_axis=None, batch_shape_meta=[B]) (construction stack None)>, | |
'energy': <_TemplateLayer(LinearLayer)(:template:linear) 'output/energy' out_type=Data(shape=(None, 1), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|1]) (construction stack 'energy_reinterpreted')>, | |
'energy_in': <_TemplateLayer(CombineLayer)(:template:combine) 'output/energy_in' out_type=Data(shape=(None, 1024), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|1024]) (construction stack 'energy_tanh')>, | |
'energy_reinterpreted': <_TemplateLayer(ReinterpretDataLayer)(:template:reinterpret_data) 'output/energy_reinterpreted' out_type=Data(shape=(None, 1), batch_shape_meta=[B,T|'spatial:0:lstm0_pool',F|1]) (construction stack 'att_weights')>, | |
'energy_tanh': <_TemplateLayer(ActivationLayer)(:template:activation) 'output/energy_tanh' out_type=Data(shape=(None, 1024), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|1024]) (construction stack 'energy')>, | |
'output': <_TemplateLayer(ChoiceLayer)(:template:choice) 'output/output' out_type=Data(shape=(), dtype='int32', sparse=True, dim=10025, time_dim_axis=None, batch_shape_meta=[B]) (construction stack None)>, | |
'output_prob': <_TemplateLayer(SoftmaxLayer)(:template:softmax) 'output/output_prob' out_type=Data(shape=(10025,), time_dim_axis=None, batch_shape_meta=[B,F|10025]) (construction stack None)>, | |
'p_t_in': <_TemplateLayer(EvalLayer)(:template:eval) 'output/p_t_in' out_type=Data(shape=(), time_dim_axis=None, batch_shape_meta=[B]) (construction stack 'att_weights')>, | |
'readout': <_TemplateLayer(ReduceOutLayer)(:template:reduce_out) 'output/readout' out_type=Data(shape=(500,), time_dim_axis=None, batch_shape_meta=[B,F|500]) (construction stack 'output_prob')>, | |
'readout_in': <_TemplateLayer(LinearLayer)(:template:linear) 'output/readout_in' out_type=Data(shape=(1000,), time_dim_axis=None, batch_shape_meta=[B,F|1000]) (construction stack 'readout')>, | |
's': <_TemplateLayer(RnnCellLayer)(:template:rnn_cell) 'output/s' out_type=Data(shape=(1000,), time_dim_axis=None, batch_shape_meta=[B,F|1000]) (construction stack 'readout_in')>, | |
's_transformed': <_TemplateLayer(LinearLayer)(:template:linear) 'output/s_transformed' out_type=Data(shape=(1024,), time_dim_axis=None, batch_shape_meta=[B,F|1024]) (construction stack 'energy_in')>, | |
'target_embed': <_TemplateLayer(LinearLayer)(:template:linear) 'output/target_embed' out_type=Data(shape=(621,), time_dim_axis=None, batch_shape_meta=[B,F|621]) (construction stack 's')>, | |
'weight_feedback': <_TemplateLayer(LinearLayer)(:template:linear) 'output/weight_feedback' out_type=Data(shape=(None, 1024), batch_dim_axis=1, batch_shape_meta=[T|'spatial:0:lstm0_pool',B,F|1024]) (construction stack 'energy_in')>} | |
Collected (unique) exceptions during template construction: | |
(Note that many of these can be ignored, or are expected.) | |
EXCEPTION | |
NetworkConstructionDependencyLoopException: Error: There is a dependency loop on layer 'accum_att_weights'. | |
Construction stack (most recent first): | |
accum_att_weights | |
weight_feedback | |
energy_in | |
energy_tanh | |
energy | |
energy_reinterpreted | |
att_weights | |
att0 | |
att | |
s | |
readout_in | |
readout | |
output_prob | |
EXCEPTION | |
CannotHandleUndefinedSourcesException: 's_transformed': cannot handle undefined sources without defined out_type. | |
{'activation': None, | |
'loss': None, | |
'n_out': 1024, | |
'network': <TFNetwork 'root/output:rec-subnet' parent_net=<TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>>, | |
'size_target': None, | |
'sources': [None], | |
'target': None, | |
'with_bias': False} | |
Exception creating layer root/'output' of class RecLayer with opts: | |
{'_target_layers': {}, | |
'cheating': False, | |
'max_seq_len': <tf.Tensor 'max_seq_len_encoder:0' shape=() dtype=int32>, | |
'n_out': <class 'Util.NotSpecified'>, | |
'name': 'output', | |
'network': <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>>, | |
'sources': [], | |
'target': 'classes', | |
'unit': {'accum_att_weights': {'class': 'eval', | |
'eval': 'source(0) + source(1) * source(2) * ' | |
'0.5', | |
'from': ['prev:accum_att_weights', | |
'att_weights', | |
'base:inv_fertility'], | |
'out_type': {'dim': 1, 'shape': (None, 1)}}, | |
'att': {'axes': 'except_batch', | |
'class': 'merge_dims', | |
'from': ['att0']}, | |
'att0': {'base': 'base:enc_value', | |
'class': 'generic_attention', | |
'weights': 'att_weights'}, | |
'att_weights': {'class': 'softmax_over_spatial', | |
'from': ['energy_reinterpreted'], | |
'window_size': 5, | |
'window_start': 'p_t_in'}, | |
'end': {'class': 'compare', 'from': ['output'], 'value': 0}, | |
'energy': {'activation': None, | |
'class': 'linear', | |
'from': ['energy_tanh'], | |
'n_out': 1, | |
'with_bias': False}, | |
'energy_in': {'class': 'combine', | |
'from': ['base:enc_ctx', | |
'weight_feedback', | |
's_transformed'], | |
'kind': 'add', | |
'n_out': 1024}, | |
'energy_reinterpreted': {'class': 'reinterpret_data', | |
'enforce_batch_major': True, | |
'from': 'energy', | |
'trainable': False}, | |
'energy_tanh': {'activation': 'tanh', | |
'class': 'activation', | |
'from': ['energy_in']}, | |
'output': {'beam_size': 12, | |
'cheating': False, | |
'class': 'choice', | |
'from': ['output_prob'], | |
'initial_output': 0, | |
'target': 'classes'}, | |
'output_prob': {'class': 'softmax', | |
'dropout': 0.3, | |
'from': ['readout'], | |
'loss': 'ce', | |
'loss_only_on_non_search': True, | |
'loss_opts': {'label_smoothing': 0}, | |
'target': 'classes'}, | |
'p_t': {'class': 'eval', | |
'eval': 'tf.to_float(source(0))', | |
'from': 'p_t_in'}, | |
'p_t_in': {'class': 'eval', | |
'eval': 'tf.squeeze(tf.argmax(source(0), axis=1, ' | |
'output_type=tf.int32), axis=1)', | |
'from': 'prev:att_weights', | |
'out_type': {'batch_dim_axis': 0, | |
'dtype': 'float32', | |
'shape': ()}}, | |
'readout': {'class': 'reduce_out', | |
'from': ['readout_in'], | |
'mode': 'max', | |
'num_pieces': 2}, | |
'readout_in': {'activation': None, | |
'class': 'linear', | |
'from': ['s', 'prev:target_embed', 'att'], | |
'n_out': 1000}, | |
's': {'class': 'rnn_cell', | |
'from': ['prev:target_embed', 'prev:att'], | |
'n_out': 1000, | |
'unit': 'LSTMBlock'}, | |
's_transformed': {'activation': None, | |
'class': 'linear', | |
'from': ['s'], | |
'n_out': 1024, | |
'with_bias': False}, | |
'target_embed': {'activation': None, | |
'class': 'linear', | |
'from': ['output'], | |
'initial_output': 0, | |
'n_out': 621, | |
'with_bias': False}, | |
'weight_feedback': {'activation': None, | |
'class': 'linear', | |
'from': ['prev:accum_att_weights'], | |
'n_out': 1024, | |
'with_bias': False}}} | |
Unhandled exception <class 'AssertionError'> in thread <_MainThread(MainThread, started 140613301045056)>, proc 20215. | |
Thread current, main, <_MainThread(MainThread, started 140613301045056)>: | |
(Excluded thread.) | |
That were all threads. | |
EXCEPTION | |
Traceback (most recent call last): | |
File "./returnn/rnn.py", line 654, in <module> | |
line: main(sys.argv) | |
locals: | |
main = <local> <function main at 0x7fe30c132158> | |
sys = <local> <module 'sys' (built-in)> | |
sys.argv = <local> ['./returnn/rnn.py', 'local-heuristic.argmax.win05.exp3.ctc.config'], _[0]: {len = 16} | |
File "./returnn/rnn.py", line 642, in main | |
line: execute_main_task() | |
locals: | |
execute_main_task = <global> <function execute_main_task at 0x7fe30c132048> | |
File "./returnn/rnn.py", line 451, in execute_main_task | |
line: engine.init_train_from_config(config, train_data, dev_data, eval_data) | |
locals: | |
engine = <global> <TFEngine.Engine object at 0x7fe04dbe8748> | |
engine.init_train_from_config = <global> <bound method Engine.init_train_from_config of <TFEngine.Engine object at 0x7fe04dbe8748>> | |
config = <global> <Config.Config object at 0x7fe3147c69e8> | |
train_data = <global> <LibriSpeechCorpus 'train' epoch=1> | |
dev_data = <global> <LibriSpeechCorpus 'dev' epoch=1> | |
eval_data = <global> None | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFEngine.py", line 891, in init_train_from_config | |
line: self.init_network_from_config(config) | |
locals: | |
self = <local> <TFEngine.Engine object at 0x7fe04dbe8748> | |
self.init_network_from_config = <local> <bound method Engine.init_network_from_config of <TFEngine.Engine object at 0x7fe04dbe8748>> | |
config = <local> <Config.Config object at 0x7fe3147c69e8> | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFEngine.py", line 934, in init_network_from_config | |
line: self._init_network(net_desc=net_dict, epoch=self.epoch) | |
locals: | |
self = <local> <TFEngine.Engine object at 0x7fe04dbe8748> | |
self._init_network = <local> <bound method Engine._init_network of <TFEngine.Engine object at 0x7fe04dbe8748>> | |
net_desc = <not found> | |
net_dict = <local> {'source': {'class': 'eval', 'eval': 'tf.clip_by_value(source(0), -3.0, 3.0)'}, 'lstm0_fw': {'class': 'rec', 'unit': 'nativelstm2', 'n_out': 1024, 'direction': 1, 'from': ['source']}, 'lstm0_bw': {'class': 'rec', 'unit': 'nativelstm2', 'n_out': 1024, 'direction': -1, 'from': ['source']}, 'lstm0_p..., len = 14 | |
epoch = <local> None | |
self.epoch = <local> 1 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFEngine.py", line 1081, in _init_network | |
line: self.network, self.updater = self.create_network( | |
config=self.config, | |
rnd_seed=net_random_seed, | |
train_flag=train_flag, eval_flag=self.use_eval_flag, search_flag=self.use_search_flag, | |
initial_learning_rate=getattr(self, "initial_learning_rate", None), | |
net_dict=net_desc) | |
locals: | |
self = <local> <TFEngine.Engine object at 0x7fe04dbe8748> | |
self.network = <local> None | |
self.updater = <local> None | |
self.create_network = <local> <bound method Engine.create_network of <class 'TFEngine.Engine'>> | |
config = <not found> | |
self.config = <local> <Config.Config object at 0x7fe3147c69e8> | |
rnd_seed = <not found> | |
net_random_seed = <local> 1 | |
train_flag = <local> <tf.Tensor 'globals/train_flag:0' shape=() dtype=bool> | |
eval_flag = <not found> | |
self.use_eval_flag = <local> True | |
search_flag = <not found> | |
self.use_search_flag = <local> False | |
initial_learning_rate = <not found> | |
getattr = <builtin> <built-in function getattr> | |
net_dict = <not found> | |
net_desc = <local> {'source': {'class': 'eval', 'eval': 'tf.clip_by_value(source(0), -3.0, 3.0)'}, 'lstm0_fw': {'class': 'rec', 'unit': 'nativelstm2', 'n_out': 1024, 'direction': 1, 'from': ['source']}, 'lstm0_bw': {'class': 'rec', 'unit': 'nativelstm2', 'n_out': 1024, 'direction': -1, 'from': ['source']}, 'lstm0_p..., len = 14 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFEngine.py", line 1113, in create_network | |
line: network.construct_from_dict(net_dict) | |
locals: | |
network = <local> <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> | |
network.construct_from_dict = <local> <bound method TFNetwork.construct_from_dict of <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>>> | |
net_dict = <local> {'source': {'class': 'eval', 'eval': 'tf.clip_by_value(source(0), -3.0, 3.0)'}, 'lstm0_fw': {'class': 'rec', 'unit': 'nativelstm2', 'n_out': 1024, 'direction': 1, 'from': ['source']}, 'lstm0_bw': {'class': 'rec', 'unit': 'nativelstm2', 'n_out': 1024, 'direction': -1, 'from': ['source']}, 'lstm0_p..., len = 14 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetwork.py", line 460, in construct_from_dict | |
line: self.construct_layer(net_dict, name) | |
locals: | |
self = <local> <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> | |
self.construct_layer = <local> <bound method TFNetwork.construct_layer of <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>>> | |
net_dict = <local> {'source': {'class': 'eval', 'eval': 'tf.clip_by_value(source(0), -3.0, 3.0)'}, 'lstm0_fw': {'class': 'rec', 'unit': 'nativelstm2', 'n_out': 1024, 'direction': 1, 'from': ['source']}, 'lstm0_bw': {'class': 'rec', 'unit': 'nativelstm2', 'n_out': 1024, 'direction': -1, 'from': ['source']}, 'lstm0_p..., len = 14 | |
name = <local> 'decision', len = 8 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetwork.py", line 652, in construct_layer | |
line: layer_class.transform_config_dict(layer_desc, network=self, get_layer=get_layer) | |
locals: | |
layer_class = <local> <class 'TFNetworkRecLayer.DecideLayer'> | |
layer_class.transform_config_dict = <local> <bound method BaseChoiceLayer.transform_config_dict of <class 'TFNetworkRecLayer.DecideLayer'>> | |
layer_desc = <local> {'loss': 'edit_distance', 'target': 'classes', 'loss_only_on_non_search': False} | |
network = <not found> | |
self = <local> <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> | |
get_layer = <local> <function TFNetwork.construct_layer.<locals>.get_layer at 0x7fe3147289d8> | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkRecLayer.py", line 4089, in transform_config_dict | |
line: super(BaseChoiceLayer, cls).transform_config_dict(d, network=network, get_layer=get_layer) | |
locals: | |
super = <builtin> <class 'super'> | |
BaseChoiceLayer = <global> <class 'TFNetworkRecLayer.BaseChoiceLayer'> | |
cls = <local> <class 'TFNetworkRecLayer.DecideLayer'> | |
transform_config_dict = <not found> | |
d = <local> {'loss': 'edit_distance', 'target': 'classes', 'loss_only_on_non_search': False} | |
network = <local> <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> | |
get_layer = <local> <function TFNetwork.construct_layer.<locals>.get_layer at 0x7fe3147289d8> | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkLayer.py", line 448, in transform_config_dict | |
line: for src_name in src_names | |
locals: | |
src_name = <not found> | |
src_names = <local> ['output'], _[0]: {len = 6} | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkLayer.py", line 449, in <listcomp> | |
line: d["sources"] = [ | |
get_layer(src_name) | |
for src_name in src_names | |
if not src_name == "none"] | |
locals: | |
d = <not found> | |
get_layer = <local> <function TFNetwork.construct_layer.<locals>.get_layer at 0x7fe3147289d8> | |
src_name = <local> 'output', len = 6 | |
src_names = <not found> | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetwork.py", line 607, in get_layer | |
line: return self.construct_layer(net_dict=net_dict, name=src_name) # set get_layer to wrap construct_layer | |
locals: | |
self = <local> <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> | |
self.construct_layer = <local> <bound method TFNetwork.construct_layer of <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>>> | |
net_dict = <local> {'source': {'class': 'eval', 'eval': 'tf.clip_by_value(source(0), -3.0, 3.0)'}, 'lstm0_fw': {'class': 'rec', 'unit': 'nativelstm2', 'n_out': 1024, 'direction': 1, 'from': ['source']}, 'lstm0_bw': {'class': 'rec', 'unit': 'nativelstm2', 'n_out': 1024, 'direction': -1, 'from': ['source']}, 'lstm0_p..., len = 14 | |
name = <not found> | |
src_name = <local> 'output', len = 6 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetwork.py", line 655, in construct_layer | |
line: return add_layer(name=name, layer_class=layer_class, **layer_desc) | |
locals: | |
add_layer = <local> <bound method TFNetwork.add_layer of <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>>> | |
name = <local> 'output', len = 6 | |
layer_class = <local> <class 'TFNetworkRecLayer.RecLayer'> | |
layer_desc = <local> {'cheating': False, 'unit': {'output': {'class': 'choice', 'target': 'classes', 'beam_size': 12, 'cheating': False, 'from': ['output_prob'], 'initial_output': 0}, 'end': {'class': 'compare', 'from': ['output'], 'value': 0}, 'target_embed': {'class': 'linear', 'activation': None, 'with_bias': Fals..., len = 7 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetwork.py", line 760, in add_layer | |
line: layer = self._create_layer(name=name, layer_class=layer_class, **layer_desc) | |
locals: | |
layer = <not found> | |
self = <local> <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> | |
self._create_layer = <local> <bound method TFNetwork._create_layer of <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>>> | |
name = <local> 'output', len = 6 | |
layer_class = <local> <class 'TFNetworkRecLayer.RecLayer'> | |
layer_desc = <local> {'cheating': False, 'unit': {'output': {'class': 'choice', 'target': 'classes', 'beam_size': 12, 'cheating': False, 'from': ['output_prob'], 'initial_output': 0}, 'end': {'class': 'compare', 'from': ['output'], 'value': 0}, 'target_embed': {'class': 'linear', 'activation': None, 'with_bias': Fals..., len = 7 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetwork.py", line 701, in _create_layer | |
line: layer_desc["output"] = layer_class.get_out_data_from_opts(**layer_desc) | |
locals: | |
layer_desc = <local> {'cheating': False, 'unit': {'output': {'class': 'choice', 'target': 'classes', 'beam_size': 12, 'cheating': False, 'from': ['output_prob'], 'initial_output': 0}, 'end': {'class': 'compare', 'from': ['output'], 'value': 0}, 'target_embed': {'class': 'linear', 'activation': None, 'with_bias': Fals..., len = 9 | |
layer_class = <local> <class 'TFNetworkRecLayer.RecLayer'> | |
layer_class.get_out_data_from_opts = <local> <bound method RecLayer.get_out_data_from_opts of <class 'TFNetworkRecLayer.RecLayer'>> | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkRecLayer.py", line 362, in get_out_data_from_opts | |
line: subnet = _SubnetworkRecCell( | |
parent_net=kwargs["network"], net_dict=unit, source_data=source_data, rec_layer_name=kwargs["name"]) | |
locals: | |
subnet = <not found> | |
_SubnetworkRecCell = <global> <class 'TFNetworkRecLayer._SubnetworkRecCell'> | |
parent_net = <not found> | |
kwargs = <local> {'cheating': False, 'target': 'classes', 'max_seq_len': <tf.Tensor 'max_seq_len_encoder:0' shape=() dtype=int32>, 'n_out': <class 'Util.NotSpecified'>, '_target_layers': {}, 'name': 'output', 'network': <TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>>}, len = 7 | |
net_dict = <not found> | |
unit = <local> {'output': {'class': 'choice', 'target': 'classes', 'beam_size': 12, 'cheating': False, 'from': ['output_prob'], 'initial_output': 0}, 'end': {'class': 'compare', 'from': ['output'], 'value': 0}, 'target_embed': {'class': 'linear', 'activation': None, 'with_bias': False, 'from': ['output'], 'n_ou..., len = 19 | |
source_data = <local> None | |
rec_layer_name = <not found> | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkRecLayer.py", line 917, in __init__ | |
line: self._construct_template() | |
locals: | |
self = <local> <_SubnetworkRecCell of None> | |
self._construct_template = <local> <bound method _SubnetworkRecCell._construct_template of <_SubnetworkRecCell of None>> | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkRecLayer.py", line 1229, in _construct_template | |
line: direct_get_layer.construct(layer.name) | |
locals: | |
direct_get_layer = <local> <RecLayer construct template GetLayer>(safe False, allow_construct_in_call_nrs None, allow_uninitialized_template False, count 1, parents None) | |
direct_get_layer.construct = <local> <bound method _SubnetworkRecCell._construct_template.<locals>.GetLayer.construct of <RecLayer construct template GetLayer>(safe False, allow_construct_in_call_nrs None, allow_uninitialized_template False, count 1, parents None)> | |
layer = <local> <_TemplateLayer(EvalLayer)(:template:eval) 'output/p_t_in' out_type=Data(shape=(), time_dim_axis=None, batch_shape_meta=[B]) (construction stack 'att_weights')> | |
layer.name = <local> 'p_t_in', len = 6 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkRecLayer.py", line 1021, in construct | |
line: self.__call__(layer_name_) | |
locals: | |
self = <local> <RecLayer construct template GetLayer>(safe False, allow_construct_in_call_nrs None, allow_uninitialized_template False, count 1, parents None) | |
self.__call__ = <local> <bound method _SubnetworkRecCell._construct_template.<locals>.GetLayer.__call__ of <RecLayer construct template GetLayer>(safe False, allow_construct_in_call_nrs None, allow_uninitialized_template False, count 1, parents None)> | |
layer_name_ = <local> 'p_t_in', len = 6 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkRecLayer.py", line 1188, in __call__ | |
line: self.net.construct_layer( | |
net_dict=self.net_dict, name=name, | |
get_layer=default_get_layer, add_layer=default_get_layer.add_templated_layer) | |
locals: | |
self = <local> <_SubnetworkRecCell of None> | |
self.net = <local> <TFNetwork 'root/output:rec-subnet' parent_net=<TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> | |
self.net.construct_layer = <local> <bound method TFNetwork.construct_layer of <TFNetwork 'root/output:rec-subnet' parent_net=<TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>>> | |
net_dict = <not found> | |
self.net_dict = <local> {'output': {'class': 'choice', 'target': 'classes', 'beam_size': 12, 'cheating': False, 'from': ['output_prob'], 'initial_output': 0}, 'end': {'class': 'compare', 'from': ['output'], 'value': 0}, 'target_embed': {'class': 'linear', 'activation': None, 'with_bias': False, 'from': ['output'], 'n_ou..., len = 19 | |
name = <local> 'p_t_in', len = 6 | |
get_layer = <not found> | |
default_get_layer = <local> <RecLayer construct template GetLayer>(safe False, allow_construct_in_call_nrs None, allow_uninitialized_template False, count 0, parents 'p_t_in') | |
add_layer = <not found> | |
default_get_layer.add_templated_layer = <local> <bound method _SubnetworkRecCell._construct_template.<locals>.GetLayer.add_templated_layer of <RecLayer construct template GetLayer>(safe False, allow_construct_in_call_nrs None, allow_uninitialized_template False, count 0, parents 'p_t_in')> | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetwork.py", line 655, in construct_layer | |
line: return add_layer(name=name, layer_class=layer_class, **layer_desc) | |
locals: | |
add_layer = <local> <bound method _SubnetworkRecCell._construct_template.<locals>.GetLayer.add_templated_layer of <RecLayer construct template GetLayer>(safe False, allow_construct_in_call_nrs None, allow_uninitialized_template False, count 0, parents 'p_t_in')> | |
name = <local> 'p_t_in', len = 6 | |
layer_class = <local> <class 'TFNetworkLayer.EvalLayer'> | |
layer_desc = <local> {'eval': 'tf.squeeze(tf.argmax(source(0), axis=1, output_type=tf.int32), axis=1)', 'out_type': {'shape': (), 'batch_dim_axis': 0, 'dtype': 'float32'}, 'sources': [<_TemplateLayer(SoftmaxOverSpatialLayer)(:prev:softmax_over_spatial) 'output/prev:att_weights' out_type=Data(shape=(1, None), time_dim... | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkRecLayer.py", line 1041, in add_templated_layer | |
line: output = layer_class.get_out_data_from_opts(**layer_desc) | |
locals: | |
output = <not found> | |
layer_class = <local> <class 'TFNetworkLayer.EvalLayer'> | |
layer_class.get_out_data_from_opts = <local> <bound method CombineLayer.get_out_data_from_opts of <class 'TFNetworkLayer.EvalLayer'>> | |
layer_desc = <local> {'eval': 'tf.squeeze(tf.argmax(source(0), axis=1, output_type=tf.int32), axis=1)', 'out_type': {'shape': (), 'batch_dim_axis': 0, 'dtype': 'float32'}, 'sources': [<_TemplateLayer(SoftmaxOverSpatialLayer)(:prev:softmax_over_spatial) 'output/prev:att_weights' out_type=Data(shape=(1, None), time_dim... | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkLayer.py", line 6069, in get_out_data_from_opts | |
line: return super(CombineLayer, cls).get_out_data_from_opts(n_out=n_out, out_type=out_type_, sources=sources, **kwargs) | |
locals: | |
super = <builtin> <class 'super'> | |
CombineLayer = <global> <class 'TFNetworkLayer.CombineLayer'> | |
cls = <local> <class 'TFNetworkLayer.EvalLayer'> | |
get_out_data_from_opts = <not found> | |
n_out = <local> <class 'Util.NotSpecified'> | |
out_type = <local> {'shape': (), 'batch_dim_axis': 0, 'dtype': 'float32'} | |
out_type_ = <local> {'name': 'p_t_in_output', 'shape': (), 'dtype': 'float32', 'sparse': False, 'dim': 1, 'batch_dim_axis': 0, 'time_dim_axis': 2, 'feature_dim_axis': 1}, len = 8 | |
sources = <local> [<_TemplateLayer(SoftmaxOverSpatialLayer)(:prev:softmax_over_spatial) 'output/prev:att_weights' out_type=Data(shape=(1, None), time_dim_axis=2, feature_dim_axis=1, batch_shape_meta=[B,F|1,T|'spatial:0:lstm0_pool']) (construction stack None)>] | |
kwargs = <local> {'eval': 'tf.squeeze(tf.argmax(source(0), axis=1, output_type=tf.int32), axis=1)', 'name': 'p_t_in', 'network': <TFNetwork 'root/output:rec-subnet' parent_net=<TFNetwork 'root' train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=bool>> train=<tf.Tensor 'globals/train_flag:0' shape=() dtype=boo... | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkLayer.py", line 227, in get_out_data_from_opts | |
line: return cls._base_get_out_data_from_opts(**kwargs) | |
locals: | |
cls = <local> <class 'TFNetworkLayer.EvalLayer'> | |
cls._base_get_out_data_from_opts = <local> <bound method LayerBase._base_get_out_data_from_opts of <class 'TFNetworkLayer.EvalLayer'>> | |
kwargs = <local> {'n_out': <class 'Util.NotSpecified'>, 'out_type': {'name': 'p_t_in_output', 'shape': (), 'dtype': 'float32', 'sparse': False, 'dim': 1, 'batch_dim_axis': 0, 'time_dim_axis': 2, 'feature_dim_axis': 1}, 'sources': [<_TemplateLayer(SoftmaxOverSpatialLayer)(:prev:softmax_over_spatial) 'output/prev:a..., len = 6 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFNetworkLayer.py", line 324, in _base_get_out_data_from_opts | |
line: output = Data(**out_type) | |
locals: | |
output = <not found> | |
Data = <global> <class 'TFUtil.Data'> | |
out_type = <local> {'name': 'p_t_in_output', 'shape': (), 'dtype': 'float32', 'sparse': False, 'dim': 1, 'batch_dim_axis': 0, 'time_dim_axis': 2, 'feature_dim_axis': 1, 'beam': None}, len = 9 | |
File "/home/ubuntu/rwth-i6/returnn-experiments/2018-asr-attention/librispeech/full-setup-attention/returnn/TFUtil.py", line 559, in __init__ | |
line: assert 0 <= feature_dim_axis < self.batch_ndim | |
locals: | |
feature_dim_axis = <local> 1 | |
self = <local> !AttributeError: 'Data' object has no attribute 'time_dim_axis' | |
self.batch_ndim = <local> 1 | |
AssertionError |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment