diff options
author | 2017-02-06 21:48:38 -0800 | |
---|---|---|
committer | 2017-02-06 22:14:44 -0800 | |
commit | 8b5a8df5e3a116510fbb7448063bfaf0bb809e65 (patch) | |
tree | 17d8ae1041c7ae8a5a50958aad2f9ab4b8aee077 | |
parent | 7a0d1b12635e9a671ed8c538f5812a8836fd6f1d (diff) |
Update generated Python Op docs.
Change: 146748564
-rw-r--r-- | tensorflow/g3doc/api_docs/python/contrib.training.md | 13 | ||||
-rw-r--r-- | tensorflow/g3doc/api_docs/python/functions_and_classes/shard4/tf.contrib.training.batch_sequences_with_states.md | 13 |
2 files changed, 22 insertions, 4 deletions
diff --git a/tensorflow/g3doc/api_docs/python/contrib.training.md b/tensorflow/g3doc/api_docs/python/contrib.training.md index 8b8847d41d..1804d5339b 100644 --- a/tensorflow/g3doc/api_docs/python/contrib.training.md +++ b/tensorflow/g3doc/api_docs/python/contrib.training.md @@ -15,7 +15,7 @@ like to store state in the forward direction across segments of an example. - - - -### `tf.contrib.training.batch_sequences_with_states(input_key, input_sequences, input_context, input_length, initial_states, num_unroll, batch_size, num_threads=3, capacity=1000, allow_small_batch=True, pad=True, name=None)` {#batch_sequences_with_states} +### `tf.contrib.training.batch_sequences_with_states(input_key, input_sequences, input_context, input_length, initial_states, num_unroll, batch_size, num_threads=3, capacity=1000, allow_small_batch=True, pad=True, make_keys_unique=False, make_keys_unique_seed=None, name=None)` {#batch_sequences_with_states} Creates batches of segments of sequential input. @@ -103,7 +103,11 @@ while True: input example. This is used to keep track of the split minibatch elements of this input. Batched keys of the current iteration are made accessible via the `key` property. The shape of `input_key` (scalar) must - be fully specified. + be fully specified. Consider setting `make_keys_unique` to True when + iterating over the same input multiple times. + + **Note**: if `make_keys_unique=False` then `input_key`s must be unique. + * <b>`input_sequences`</b>: A dict mapping string names to `Tensor` values. The values must all have matching first dimension, called `value_length`. They may vary from input to input. The remainder of the shape (other than the first @@ -158,6 +162,11 @@ while True: `num_unroll`. In that case `input_length` may be `None` and is assumed to be the length of first dimension of values in `input_sequences` (i.e. `value_length`). +* <b>`make_keys_unique`</b>: Whether to append a random integer to the `input_key` in + an effort to make it unique. The seed can be set via + `make_keys_unique_seed`. +* <b>`make_keys_unique_seed`</b>: If `make_keys_unique=True` this fixes the seed with + which a random postfix is generated. * <b>`name`</b>: An op name string (optional). ##### Returns: diff --git a/tensorflow/g3doc/api_docs/python/functions_and_classes/shard4/tf.contrib.training.batch_sequences_with_states.md b/tensorflow/g3doc/api_docs/python/functions_and_classes/shard4/tf.contrib.training.batch_sequences_with_states.md index a59080eb30..63c3a47229 100644 --- a/tensorflow/g3doc/api_docs/python/functions_and_classes/shard4/tf.contrib.training.batch_sequences_with_states.md +++ b/tensorflow/g3doc/api_docs/python/functions_and_classes/shard4/tf.contrib.training.batch_sequences_with_states.md @@ -1,4 +1,4 @@ -### `tf.contrib.training.batch_sequences_with_states(input_key, input_sequences, input_context, input_length, initial_states, num_unroll, batch_size, num_threads=3, capacity=1000, allow_small_batch=True, pad=True, name=None)` {#batch_sequences_with_states} +### `tf.contrib.training.batch_sequences_with_states(input_key, input_sequences, input_context, input_length, initial_states, num_unroll, batch_size, num_threads=3, capacity=1000, allow_small_batch=True, pad=True, make_keys_unique=False, make_keys_unique_seed=None, name=None)` {#batch_sequences_with_states} Creates batches of segments of sequential input. @@ -86,7 +86,11 @@ while True: input example. This is used to keep track of the split minibatch elements of this input. Batched keys of the current iteration are made accessible via the `key` property. The shape of `input_key` (scalar) must - be fully specified. + be fully specified. Consider setting `make_keys_unique` to True when + iterating over the same input multiple times. + + **Note**: if `make_keys_unique=False` then `input_key`s must be unique. + * <b>`input_sequences`</b>: A dict mapping string names to `Tensor` values. The values must all have matching first dimension, called `value_length`. They may vary from input to input. The remainder of the shape (other than the first @@ -141,6 +145,11 @@ while True: `num_unroll`. In that case `input_length` may be `None` and is assumed to be the length of first dimension of values in `input_sequences` (i.e. `value_length`). +* <b>`make_keys_unique`</b>: Whether to append a random integer to the `input_key` in + an effort to make it unique. The seed can be set via + `make_keys_unique_seed`. +* <b>`make_keys_unique_seed`</b>: If `make_keys_unique=True` this fixes the seed with + which a random postfix is generated. * <b>`name`</b>: An op name string (optional). ##### Returns: |