Xiaolong Wang 82adbde878 pass layer_parameter shape to ps builder if cannot inferred from initializer ... Summary: Feed team uses distributed training and wants to also use transfer learning. Currently, transfer learning implements by overwriting the layer parameter initializer. Therefore, PS builder can't infer correctly the parameter shape. To fix this, add a field 'shape' in `layer_parameter` and set the shape if we overwrite its initializer. We also enforce the check of parameter shape between the original initializer and the loaded blob. (this adds extra cost) Differential Revision: D5520541 fbshipit-source-id: 80547dbd328b3f6cbfcea0b2daaf4004703dfe81		2017-07-31 16:04:23 -07:00
..
__init__.py	Allow to import subclasses of layers	2017-07-12 20:19:47 -07:00
add_bias.py	Add bias to cosine distance for two tower models	2017-05-25 19:50:20 -07:00
arc_cosine_feature_map.py	Fixing semi-random layer model for multi-layer models	2017-07-27 15:25:19 -07:00
batch_distill_lr_loss.py	Feature importance in dper 2.0: build network representation	2017-06-05 18:03:34 -07:00
batch_lr_loss.py	Feature importance in dper 2.0: build network representation	2017-06-05 18:03:34 -07:00
batch_mse_loss.py	Feature importance in dper 2.0: build network representation	2017-06-05 18:03:34 -07:00
batch_normalization.py	implement drelu and unittest	2017-07-20 11:50:08 -07:00
batch_sigmoid_cross_entropy_loss.py	Feature importance in dper 2.0: build network representation	2017-06-05 18:03:34 -07:00
batch_softmax_loss.py	add support for weight in batch_softmax_loss	2017-06-21 10:32:15 -07:00
build_index.py	Return top K classes	2017-07-13 00:20:00 -07:00
concat.py	dot product using matmul	2017-07-17 23:20:37 -07:00
dot_product.py	Dict fixes/improvements and unittest targets for Python 3 in caffe2 core	2017-06-29 17:05:41 -07:00
dropout.py	Adding Dropout Layer to SparseNN Model and Flow	2017-06-20 15:46:55 -07:00
fc.py
fc_without_bias.py
feature_sparse_to_dense.py	Rename SparseToDense layer	2017-06-09 12:48:27 -07:00
functional.py	make functional layer return scalar if only one output	2017-07-12 11:34:31 -07:00
gather_record.py
last_n_window_collector.py	LastNWindowCollector	2017-05-04 17:32:09 -07:00
layers.py	pass layer_parameter shape to ps builder if cannot inferred from initializer	2017-07-31 16:04:23 -07:00
pairwise_dot_product.py	BatchGatherOp	2017-07-27 10:17:42 -07:00
position_weighted.py	add truncation for sparse feature	2017-06-13 10:46:53 -07:00
random_fourier_features.py	Simplifying Random Fourier Features and layer test	2017-07-11 00:40:53 -07:00
sampling_train.py	Modify samplingTrain layer to take more general inputs	2017-07-08 22:19:55 -07:00
sampling_trainable_mixin.py
select_record_by_context.py	JoinContext	2017-05-02 17:32:26 -07:00
semi_random_features.py	Fixing semi-random layer model for multi-layer models	2017-07-27 15:25:19 -07:00
sparse_feature_hash.py	IndexHash	2017-07-07 23:06:11 -07:00
sparse_lookup.py	make SparseLookup support None pooling	2017-07-18 16:39:55 -07:00
split.py
tags.py	offline_all_gpu_experiment	2017-06-27 23:09:54 -07:00
uniform_sampling.py	Re-apply #266	2017-04-25 21:17:04 -07:00