Muller Helen

implicit val session = TestSession.createTestSession(TestSession.Mode.EAGER) // EagerSession.create() implicit val tf = session.getTF // Ops.create(session).withName("test") implicit val scope = **tf.scope()* val weightInputGate = tf.variable(getWeightMatrix(weightGatesDims, scope)) the console error : Op VariableV2...

How to update the weight about BlockLSTM ?

1

HI : now I use BlockLSTM for build lstm layer ,but I don't know how to update the lstm weight parameters , if need use blockLSTMGrad or something to do...

Could we supply the method that convert spark sql dataframe to tensorflow-java ByteNdArray object ?

2

Hi : Now I etl the raw data use spark sql generate the dataframe , I want to convert the spark dataframe object to ByteNdArray object for tensorflow model input

HI : for some etl the data, we need thses method ,but in tensorflow-java ,I can't find them

HI : I want to use some time series model ,I know we have // LSTMBlockCell.create() // GRUBlockCell.create() // BlockLSTM.create(), but I don't know how to use it, would you...

Exception in thread "main" org.tensorflow.exceptions.TensorFlowException: No gradient defined for op: SelectV2.

1

env : tensorflow-java : 0.5.0-SNAPSHOT scala :2.12.11 java : jdk 15 Hi : when I defined a tf Variable named embeddings ,and want to do some operation with it, invoke...

HI : I want to use tf.repeat ,but in tensorflow -java ，I am not found this method ,It is a import logic method for some algorithm. please

How to represent the [variable_scope] and [get_variable] in tensorflow java ?

2

HI : write the tensorflow code in python ,we always use with the variable_scope and get_variable method , like this ``` with tf.compat.v1.variable_scope('Qifficulty'): question_difficulty_weights = tf.compat.v1.get_variable('weights', [q.get_shape()[-1], 1], initializer= tf.initializers.glorot_uniform(seed=0))...

Sorry for one more help :Eager to support the Embedding layer and embedding_lookup method make real

3

HI : I want to build recommend system use deepfm model ,but it need a lots of layer ,most important layer is Embedding layer, and so many NLP work model...

Do we support tf.nn.bidirectional_dynamic_rnn ?

3

Hi: I only see CudnnRNN ,the project has not testcase for it,I not know if we support tf.nn.bidirectional_dynamic_rnn or normal simple rnn? thank you