Frank Jiang b827a40880 Implement bucket-based attention pooling for IdScoreList features (#13004) ... Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/13004 Implement BucketWeighted model layer, which learns a weight for each possible score in an IdScoreList. Here, we assume that the scores in the IdScoreList have already been converted into the appropriate 'buckets'. If this is not done, then essentially each score represents its own bucket. We assume that the scores/buckets are integers, and if max_score is not set, we assume that the maximum cardinality of the score is less than or equal to the cardinality of the ids. Reviewed By: chonglinsun Differential Revision: D10413186 fbshipit-source-id: 743e643a1b36adf124502a8b6b29976158cdb130		2018-10-25 18:04:08 -07:00
..
__init__.py
adaptive_weight.py
add_bias.py
arc_cosine_feature_map.py
batch_distill_lr_loss.py
batch_lr_loss.py
batch_mse_loss.py
batch_normalization.py
batch_sigmoid_cross_entropy_loss.py
batch_softmax_loss.py
blob_weighted_sum.py
bucket_weighted.py	Implement bucket-based attention pooling for IdScoreList features (#13004)	2018-10-25 18:04:08 -07:00
build_index.py
concat.py	refactor preproc, support dense in TumHistory layer	2018-09-05 16:10:13 -07:00
constant_weight.py
conv.py
dropout.py
fc.py
fc_without_bias.py
feature_sparse_to_dense.py
functional.py
gather_record.py
homotopy_weight.py
label_smooth.py
last_n_window_collector.py
layer_normalization.py	Allow use substitute ops for LayerNorm (#12177)	2018-10-11 17:36:10 -07:00
layers.py
margin_rank_loss.py
merge_id_lists.py
pairwise_similarity.py	move matrix formation for dot products to precompute/request-only (#10531)	2018-08-15 11:02:10 -07:00
position_weighted.py	Remove unused code base for distributed training (#10282)	2018-08-16 20:10:17 -07:00
random_fourier_features.py
reservoir_sampling.py
sampling_train.py
sampling_trainable_mixin.py
select_record_by_context.py
semi_random_features.py
sparse_feature_hash.py
sparse_lookup.py	Revert D10123465: Set the correct engine name for position weighted pooling when fp16 is used for training	2018-10-15 01:53:48 -07:00
split.py
tags.py	Remove unused code base for distributed training (#10282)	2018-08-16 20:10:17 -07:00
uniform_sampling.py