Tristan Rice 676fc929b7 [caffe2] fix type and shape inference for common gradient ops (#35857) ...

Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/35857

This fixes a lot of common ops for InferBlobShapesAndTypes as well as adds support for testing the inferred shapes and types of gradient ops.

Ops:
* Concat
* Split
* LeakyReLU
* Relu
* Prelu
* Gelu
* Elu
* Sinh, Tanh, Cosh
* Abs
* ... and a number of other simple element wise ops

Test Plan:
Added support to hypothesis test to check the shape and type of gradient ops.

Enabled it for all the ops I fixed the shape and type inference for.

  buck test caffe2/caffe2/python/operator_test:

Reviewed By: pradeepd24

Differential Revision: D20806284

fbshipit-source-id: 77f796d9ff208e09e871bdbadf9a0a7c196b77f2

2020-04-02 11:17:04 -07:00

..

__init__.py

…

activation_ops_test.py

[caffe2] fix type and shape inference for common gradient ops (#35857)

2020-04-02 11:17:04 -07:00

adadelta_test.py

…

adagrad_test_helper.py

…

adagrad_test.py

…

adam_test.py

Revert D20289209: Support RowWiseSparseAdam on GPU

2020-03-18 07:35:07 -07:00

affine_channel_op_test.py

…

alias_with_name_test.py

move AliasWithNameOp to caffe2/operators

2019-12-17 02:39:40 -08:00

apmeter_test.py

…

arg_ops_test.py

…

assert_test.py

…

atomic_ops_test.py

Add 64bit atomic fetch add (#32354)

2020-01-17 11:43:43 -08:00

basic_rnn_test.py

…

batch_box_cox_test.py

…

batch_bucketize_op_test.py

Fix typos, via a Levenshtein-type corrector (#31523)

2020-01-17 16:03:19 -08:00

batch_moments_op_test.py

…

batch_sparse_to_dense_op_test.py

Add cuda version for operators BatchSparseToDense and BatchDenseToSparse (#29166)

2019-11-05 13:06:23 -08:00

bbox_transform_test.py

…

bisect_percentile_op_test.py

…

blobs_queue_db_test.py

…

boolean_mask_test.py

…

boolean_unmask_test.py

…

box_with_nms_limit_op_test.py

Fix typos, via a Levenshtein-type corrector (#31523)

2020-01-17 16:03:19 -08:00

bucketize_op_test.py

…

cast_op_test.py

…

ceil_op_test.py

…

channel_backprop_stats_op_test.py

…

channel_shuffle_test.py

batch size 0 support in ChannelShuffle DNNLOWP op (#26858)

2019-09-26 00:40:07 -07:00

channel_stats_op_test.py

…

checkpoint_test.py

…

clip_op_test.py

…

clip_tensor_op_test.py

…

collect_and_distribute_fpn_rpn_proposals_op_test.py

…

concat_split_op_test.py

[caffe2] fix type and shape inference for common gradient ops (#35857)

2020-04-02 11:17:04 -07:00

conditional_test.py

…

conftest.py

…

conv_test.py

[ROCm] Enable 3D convolutions through ROCm (#33067)

2020-02-14 13:19:10 -08:00

conv_transpose_test.py

fix zero-batch handling in convtranspose (#24341)

2019-12-18 15:06:36 -08:00

copy_ops_test.py

…

copy_rows_to_tensor_op_test.py

…

cosine_embedding_criterion_op_test.py

…

counter_ops_test.py

…

crf_test.py

[Python] Deprecate use of scipy.misc.logsumexp and scipy.misc.comb (#32209)

2020-01-15 10:40:47 -08:00

cross_entropy_ops_test.py

…

ctc_beam_search_decoder_op_test.py

…

ctc_greedy_decoder_op_test.py

…

cudnn_recurrent_test.py

…

data_couple_op_test.py

…

dataset_ops_test.py

Fix typos (#30606)

2019-12-02 20:17:42 -08:00

deform_conv_test.py

…

dense_vector_to_id_list_op_test.py

…

depthwise_3x3_conv_test.py

…

detectron_keypoints.py

…

distance_op_test.py

…

dropout_op_test.py

…

duplicate_operands_test.py

…

elementwise_linear_op_test.py

…

elementwise_logical_ops_test.py

…

elementwise_op_broadcast_test.py

Fix typos, via a Levenshtein-type corrector (#31523)

2020-01-17 16:03:19 -08:00

elementwise_ops_test.py

[caffe2] fix type and shape inference for common gradient ops (#35857)

2020-04-02 11:17:04 -07:00

emptysample_ops_test.py

…

enforce_finite_op_test.py

…

ensure_clipped_test.py

…

ensure_cpu_output_op_test.py

…

erf_op_test.py

…

expand_op_test.py

…

fc_operator_test.py

…

feature_maps_ops_test.py

…

filler_ops_test.py

…

find_op_test.py

…

flatten_op_test.py

…

flexible_top_k_test.py

…

floor_op_test.py

…

fused_nbit_rowwise_conversion_ops_test.py

[caffe2] open source 2/4-bit SLS operators (#34903)

2020-03-17 22:55:10 -07:00

fused_nbit_rowwise_test_helper.py

[caffe2] open source 2/4-bit SLS operators (#34903)

2020-03-17 22:55:10 -07:00

fused_nbit_rowwise_test.cc

[caffe2] open source 2/4-bit SLS operators (#34903)

2020-03-17 22:55:10 -07:00

gather_ops_test.py

Fix typos, via a Levenshtein-type corrector (#31523)

2020-01-17 16:03:19 -08:00

gather_ranges_op_test.py

[fix][tiny][caffe2] Avoid triggering errors when allow ratio is 100% (#34757)

2020-03-16 11:39:32 -07:00

given_tensor_byte_string_to_uint8_fill_op_test.py

…

given_tensor_fill_op_test.py

…

glu_op_test.py

…

group_conv_test.py

…

group_norm_op_test.py

…

gru_test.py

…

heatmap_max_keypoint_op_test.py

Fix typos, via a Levenshtein-type corrector (#31523)

2020-01-17 16:03:19 -08:00

hsm_test.py

…

hyperbolic_ops_test.py

[caffe2] fix type and shape inference for common gradient ops (#35857)

2020-04-02 11:17:04 -07:00

im2col_col2im_test.py

…

image_input_op_test.py

…

index_hash_ops_test.py

Support in-place update in IndexHashOp (#30275)

2019-11-22 14:49:28 -08:00

index_ops_test.py

…

instance_norm_test.py

…

integral_image_ops_test.py

…

jsd_ops_test.py

…

key_split_ops_test.py

…

lars_test.py

…

layer_norm_op_test.py

…

leaky_relu_test.py

…

learning_rate_adaption_op_test.py

…

learning_rate_op_test.py

fix composite learning rate (#26227)

2019-09-18 17:34:17 -07:00

length_split_op_test.py

…

lengths_pad_op_test.py

…

lengths_reducer_fused_nbit_rowwise_ops_test.py

Split SparseLengthsSumSparse into SparseLengthsSumSparseLookup + SparseLengthsSum (#35507)

2020-03-30 13:33:29 -07:00

lengths_tile_op_test.py

…

lengths_top_k_ops_test.py

…

listwise_l2r_operator_test.py

add LambdaRank DCG Loss Option (#23679)

2019-08-02 11:47:46 -07:00

load_save_test.py

…

locally_connected_op_test.py

…

loss_ops_test.py

…

lpnorm_op_test.py

…

map_ops_test.py

…

margin_ranking_criterion_op_test.py

…

math_ops_test.py

[Caffe2] Fix shape inference for element-wise operators (#33431)

2020-02-25 09:03:06 -08:00

matmul_op_test.py

[caffe2] remove unnecessary np.set_printoptions and fix test errors (#32475)

2020-01-22 14:49:47 -08:00

mean_op_test.py

…

merge_id_lists_op_test.py

…

mkl_conv_op_test.py

…

mkl_packed_fc_op_test.py

…

mkl_speed_test.py

…

mod_op_test.py

…

moments_op_test.py

…

momentum_sgd_test.py

Disable flaky TestMomentumSGD.test_fp16momentum_sgd (#31369)

2019-12-17 19:16:54 -08:00

mpi_test.py

…

mul_gradient_benchmark.py

optimize MulGradient for common shapes (#19705)

2019-12-11 11:39:52 -08:00

negate_gradient_op_test.py

…

ngram_ops_test.py

…

normalize_op_test.py

…

numpy_tile_op_test.py

…

one_hot_ops_test.py

Fix typos, via a Levenshtein-type corrector (#31523)

2020-01-17 16:03:19 -08:00

onnx_while_test.py

…

order_switch_test.py

…

pack_ops_test.py

Change PackSegments to ensure consistent behavior between CPU and GPU

2019-12-30 13:31:45 -08:00

pack_rnn_sequence_op_test.py

…

pad_test.py

…

partition_ops_test.py

…

percentile_op_test.py

…

piecewise_linear_transform_test.py

…

pooling_test.py

Fix typos, via a Levenshtein-type corrector (#31523)

2020-01-17 16:03:19 -08:00

prepend_dim_test.py

…

python_op_test.py

…

quantile_test.py

[Rowwise Pruning][c2 op] Add Quantile Op (#32448)

2020-01-22 16:59:56 -08:00

rand_quantization_op_speed_test.py

…

rand_quantization_op_test.py

Disables flaky test_rand_quantization (#29463)

2019-11-08 13:51:22 -08:00

rank_loss_operator_test.py

…

rebatching_queue_test.py

…

record_queue_test.py

…

recurrent_net_executor_test.py

…

recurrent_network_test.py

Fix typos, via a Levenshtein-type corrector (#31523)

2020-01-17 16:03:19 -08:00

reduce_ops_test.py

…

reduction_ops_test.py

…

reshape_ops_test.py

…

resize_op_test.py

…

rmac_regions_op_test.py

…

rnn_cell_test.py

…

roi_align_rotated_op_test.py

…

rowwise_counter_test.py

separate op for rowwise counter (#31612)

2019-12-27 12:18:39 -08:00

scale_op_test.py

…

segment_ops_test.py

…

selu_op_test.py

…

sequence_ops_test.py

…

shape_inference_test.py

add conversion functions to embedding tables (#31083)

2020-01-08 16:56:12 -08:00

sinusoid_position_encoding_op_test.py

…

softmax_ops_test.py

Support softmax with D == 0 (#29167)

2019-11-11 00:46:10 -08:00

softplus_op_test.py

…

sparse_dropout_with_replacement_op_test.py

…

sparse_gradient_checker_test.py

…

sparse_lengths_sum_benchmark.py

[caffe2] use 2-stage EmbeddingSpMDM interface (#32271)

2020-01-22 19:05:36 -08:00

sparse_normalize_test.py

Make SparseNormalize backwards compatible (#25660)

2019-09-05 15:14:21 -07:00

sparse_ops_test.py

…

sparse_to_dense_mask_op_test.py

…

spatial_bn_op_test.py

[ROCm] Enable 3D batch norms through MIOpen (#33262)

2020-02-13 11:29:51 -08:00

specialized_segment_ops_test.py

…

square_root_divide_op_test.py

…

stats_ops_test.py

…

stats_put_ops_test.py

…

string_ops_test.py

…

text_file_reader_test.py

…

thresholded_relu_op_test.py

…

tile_op_test.py

…

top_k_test.py

…

torch_integration_test.py

[2/3] Bind Bucketize to PyTorch (#33014)

2020-02-11 23:20:10 -08:00

transpose_op_test.py

…

trigonometric_op_test.py

…

unique_ops_test.py

…

unique_uniform_fill_op_test.py

…

upsample_op_test.py

…

utility_ops_test.py

[dt] [caffe2] add/fix shape inference for StumpFunc, SliceGradient and ResizeLike (#35430)

2020-03-26 17:50:32 -07:00

video_input_op_test.py

…

weight_scale_test.py

Change weight scale test to cpu only (#35346)

2020-03-25 09:18:58 -07:00

weighted_multi_sample_test.py

…

weighted_sample_test.py

…

weighted_sum_test.py

…

wngrad_test.py

…