Can we change the column name input_types to valid_inputs and have rows in that column consist of a string listing each unique variable type present in that primitives input_types? So for Haversine, valid_inputs would be "LatLong", whereas GreaterThan would be "Datetime, Numeric, Ordinal"

Hi,

first of all thanks for you work and effort with TuriCreate.

As I already saw, you are using tensorflow for some tasks, therefore, I was wondering: is it possible to export the model as SavedModel format ?

Regards, Xavi

You are correct many of our models use TensorFlow when running on Linux or on older versions of macOS. However TuriCreate does not have the ability to export to any TensorFlow format. That being said, it's certainly something that would be possible. Is there a particular model type that you are interested in exporting to TensorFlow?

We are already tracking this issue in #2930. So I'm going to close this issue as a duplicate.

Codecov Report

Merging #1341 (d30c9bd) into main (6f7cead) will increase coverage by 0.19%. The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #1341      +/-   ##
==========================================
+ Coverage   98.38%   98.58%   +0.19%     
==========================================
  Files         135      135              
  Lines       14496    14530      +34     
==========================================
+ Hits        14262    14324      +62     
+ Misses        234      206      -28     

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 019f09d...5160408. Read the comment docs.

The function list_primitives() will also return the input types and return type of the primitive. (Closes #1184, Closes #1084)

                  name       type  dask_compatible  koalas_compatible                                        description                                        input_types return_type
74           haversine  transform            False              False  Calculates the approximate haversine distance ...                                 [LatLong, LatLong]     Numeric
75           num_words  transform             True               True  Determines the number of words in a string by ...                                  [NaturalLanguage]     Numeric
76            absolute  transform             True               True           Computes the absolute value of a number.                                          [Numeric]     Numeric
77          time_since  transform             True              False  Calculates time from a value to a specified cu...                                         [Datetime]     Numeric
78  less_than_equal_to  transform             True               True  Determines if values in one list are less than...  [[Numeric, Numeric], [Datetime, Datetime], [Or...     Boolean

Hi,

first of all thanks for you work and effort with TuriCreate.

As I already saw, you are using tensorflow for some tasks, therefore, I was wondering: is it possible to export the model as SavedModel format ?

Regards, Xavi

in fact, use dense, the model need to train more than 30000 iterations , and then it will reach the acc to 0.90 + .

def RNN(x, weights, biases):

# Prepare data shape to match `rnn` function requirements
# Current data input shape: (batch_size, timesteps, n_input)
# Required shape: 'timesteps' tensors list of shape (batch_size, n_input)

# Unstack to get a list of 'timesteps' tensors of shape (batch_size, n_input)
x = tf.unstack(x, timesteps, 1)

# Define a lstm cell with tensorflow
lstm_cell = rnn.BasicLSTMCell(num_hidden, forget_bias=1.0)

# Get lstm cell output
outputs, states = rnn.static_rnn(lstm_cell, x, dtype=tf.float32)

return tf.layers.dense(outputs[-1], num_classes, activation=None, kernel_regularizer=tf.contrib.layers.l2_regularizer(1e-2)) 
#return tf.matmul(outputs[-1], weights['out']) + biases['out']

use dense: the acc is 0.507 but use tf.matmul(outputs[-1], weights['out']) + biases['out']) is 0.906