ef1b7175d5
fix compatibility with PyTorch 1.10 ( #3454 )
2021-11-04 08:20:44 +00:00
87a4cb363a
Fix #3437 ( #3440 )
2021-11-04 08:20:22 +00:00
c7f94dbf5a
[BugFix] Avoid Memory Leak Issue in PyTorch Backend ( #3386 )
...
* try to avoid memory leak
* try to avoid memory leak
* avoid memory leak with no hope
* Revert "avoid memory leak with no hope"
This reverts commit c77befe9479f46758e744642f66dd209b50eef7d.
* no message
* Update sparse.py
* Update tensor.py
Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com >
2021-11-04 08:18:40 +00:00
a63ad14bda
[PyTorch][Bugfix] Use uint8 instead of bool in pytorch to be compatible with nightly version ( #3406 )
...
* Use uint8 instead of bool in pytorch
* Handle type aliases
* Fix syntax error
Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com >
2021-11-04 08:15:46 +00:00
223795eb82
[Distributed] Distributed heterograph training ( #3069 )
...
* support hetero RGCN.
* fix.
* simplify code.
* sample_neighbors return heterograph directly.
* avoid using to_heterogeneous.
* compute canonical etypes in advance.
* fix tests.
* fix.
* fix distributed data loader for heterograph.
* use NodeDataLoader.
* fix bugs in partitioning on heterogeneous graphs.
* fix lint.
* fix tests.
* fix.
* fix.
* fix bugs.
* fix tests.
* fix.
* enable coo for distributed.
* fix.
* fix.
* fix.
* fix.
* fix.
Co-authored-by: Ubuntu <ubuntu@ip-172-31-71-112.ec2.internal >
Co-authored-by: Zheng <dzzhen@3c22fba32af5.ant.amazon.com >
2021-11-02 09:18:15 +00:00
eca9b8ef08
allow for configuring default_dir ( #3277 )
...
* allow for configuring default_dir
* allow for using DGLDEFAULTDIR environment variable
* Update env_var.rst
Co-authored-by: Quan (Andy) Gan <coin2028@hotmail.com >
Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com >
2021-08-26 17:32:16 +00:00
92550b3b03
[perf] Remove activation cache if not required. ( #3258 )
...
* upd
* fix
* upd
2021-08-20 02:44:06 +00:00
35c898c9e9
Bug fix: change th to tf ( #3182 )
...
Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com >
2021-08-20 02:14:19 +00:00
b379dbd6c8
Optimize dist_graph/_split_even_to_part memory usage ( #3132 )
...
Co-authored-by: yujingcheng02 <yujingcheng02@meituan.com >
Co-authored-by: Quan (Andy) Gan <coin2028@hotmail.com >
2021-07-15 11:51:56 +08:00
188152b853
[Feature] Add Heterograph support on Python for builtin unary msg functions (copy_u, copy_e) ( #2989 )
...
* heterograph for binary func
* Added SDDMM support
* Added unittest
* added binary test cases
* unary mfuncs works
* Fixed lint err
* lint check and others
* link check
* fixed import *_hetero issue
* lint check
* replace torch with dgl backend
* lint cehck
* removed torch from test
* skip mxnet unittest
* skip gpu test
* Remove unused/duplicated code
* minor
* changed data structure of ndata and edata
* link check
* reorganized
* minor lint
* minor lint
* raise error for udf func
* lint check
* fix for CUDA 10.1
* add a note for future cross-type max/min reducing
* Add support CUDA < 11
* lint check
* tidied C code
* remove dummy GSDDMM_hetero backward implementation
Co-authored-by: Israt Nisa <nisisrat@amazon.com >
Co-authored-by: Minjie Wang <wmjlyjemaine@gmail.com >
Co-authored-by: Quan Gan <coin2028@hotmail.com >
2021-07-06 20:41:58 +08:00
acd21a6d60
[Feature] Support direct creation from CSR and CSC ( #3045 )
...
* csr and csc creation
* fix
* fix
* fixes to adj transpose
* fine
* raise error if indptr did not match number of nodes
* fix
* huh?
* oh
Co-authored-by: Minjie Wang <wmjlyjemaine@gmail.com >
2021-06-25 10:05:47 +08:00
17d604b5c7
[Feature] Allow using NCCL for communication in dgl.NodeEmbedding and dgl.SparseOptimizer ( #2824 )
...
* Split from NCCL PR
* Fix type in comment
* Expand documentation for sparse_all_to_all_push
* Restore previous behavior in example
* Re-work optimizer to use NCCL based on gradient location
* Allow for running with embedding on CPU but using NCCL for gradient exchange
* Optimize single partition case
* Fix pylint errors
* Add missing include
* fix gradient indexing
* Fix line continuation
* Migrate 'first_step'
* Skip tests without enough GPUs to run NCCL
* Improve empty tensor handling for pytorch 1.5
* Fix indentation
* Allow multiple NCCL communicator to coexist
* Improve handling of empty message
* Update python/dgl/nn/pytorch/sparse_emb.py
Co-authored-by: xiang song(charlie.song) <classicxsong@gmail.com >
* Update python/dgl/nn/pytorch/sparse_emb.py
Co-authored-by: xiang song(charlie.song) <classicxsong@gmail.com >
* Keepy empty tensor dimensionaless
* th.empty -> th.tensor
* Preserve shape for empty non-zero dimension tensors
* Use shared state, when embedding is shared
* Add support for gathering an embedding
* Fix typo
* Fix more typos
* Fix backend call
* Use NodeDataLoader to take advantage of ddp
* Update training script to share memory
* Only squeeze last dimension
* Better handle empty message
* Keep embedding on the target device GPU if dgl_sparse if false in RGCN example
* Fix typo in comment
* Add asserts
* Improve documentation in example
Co-authored-by: xiang song(charlie.song) <classicxsong@gmail.com >
2021-06-10 21:19:00 -07:00
657c220de7
[Feature] Python interface for adjacency matrix summation and multiplication ( #2893 )
...
* test commit
* fixes
* oops
* add docs
* lint
* why does it say I have a trailing whitespace
* oh ok
* fixes
* why there's an invalid argument error
* address comments
* fix
* address comments
2021-05-17 19:47:02 +08:00
2a3a9a0011
[Bugfix] to() method ignores PyTorch default CUDA device ( #2897 )
2021-05-08 13:23:30 +08:00
e70138bbe4
[Performance][GPU] Enable GPU uniform edge sampling ( #2716 )
...
* Start on uniform GPU sampling
* Save more work
* Get cu file compiling
* Update sampling
* More changes
* Get GPU sampling for uniform probabilities solved
* Fix batch tensor migration
* Fix
* update kernels
* expand blocking
* Undo testing change
* Cut down on sampling overhead
* Fix replacement
* Update unit tests
* Add option to gpu sample in graphsage
* Copy only csc to gpu
* Add ogbn support
* Fix linting
* Remove nvtx from sample
* Improve documentation and error checking
* Expand documentation
* Update assert checking
* delete extra space
* Use standard dataloader when dataset is a dictionary
* ogb -> ogbn
* Fix edge selection determinism
* Fix typos
* Remove nvtx
* Add comment for self.fanout_arrays and assert
* Fix linting
* Migrate to scalarbatcher
* Fix indentation
* Fix batcher
* Fix indexing
* Only use databatcher for GPU
* Convert to DGL NDArray to PyTorch Tensor
* Add optimization for PyTorch's F.tensor() for list of GPU tensors
Co-authored-by: Da Zheng <zhengda1936@gmail.com >
2021-04-14 18:55:09 -07:00
e6bf54cdcb
[Feature] Support aggregate multiple edge features in to_simple. ( #2623 )
...
* upd
* fix
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* fix
* refactor
* upd test
* large feat_len or n in segment reduce
* lint
2021-02-21 10:05:42 +08:00
cf8a3fb305
[hotfix] Address the performance issue of topk. ( #2628 )
...
* upd
* fix
* upd
* upd
* udp
* upd
* upd
* upd
* upd
* upd
* upd
Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com >
2021-02-11 00:00:39 +08:00
fb4a0508c6
[bugfix] Solve the boundary issue in backward function of segment sum ( #2610 )
...
* upd
* trigger
* upd
2021-02-03 18:50:07 +08:00
fb3c07094f
Revert part of #2563 ( #2584 )
2021-01-28 14:53:55 +08:00
878acdb0fc
Revert "Refactor code for retaining formats in message-passing. ( #2570 )" ( #2583 )
...
This reverts commit a613ad8806
2021-01-28 14:04:07 +08:00
7bab1365e2
[feature] Supporting half precision floating data type (fp16). ( #2552 )
...
* add tvm as submodule
* compilation is ok but calling fails
* can call now
* pack multiple modules, change names
* upd
* upd
* upd
* fix cmake
* upd
* upd
* upd
* upd
* fix
* relative path
* upd
* upd
* upd
* singleton
* upd
* trigger
* fix
* upd
* count reducible
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* only keep related files
* upd
* upd
* upd
* upd
* lint
* lint
* lint
* lint
* pylint
* upd
* upd
* compilation
* fix
* upd
* upd
* upd
* upd
* upd
* upd
* upd doc
* refactor
* fix
* upd number
Co-authored-by: Zhi Lin <linzhilynn@gmail.com >
Co-authored-by: Ubuntu <ubuntu@ip-172-31-42-78.us-east-2.compute.internal >
Co-authored-by: Ubuntu <ubuntu@ip-172-31-21-156.us-east-2.compute.internal >
Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com >
2021-01-28 11:21:58 +08:00
a7e941c379
[Feature] Add support for sparse embedding ( #2451 )
...
* Add sparse embedding for dgl and update rgcn example
* upd
* Fix
* Revert "Fix"
This reverts commit 4da87cdfb8b8c3506b7fc7376cd2385ba8045c2a.
* Fix
* upd
* upd
* Fix
* Add unitest and update impl
* fix
* Clean up rgcn example code
* upd
* upd
* update
* Fix
* update score
* sparse for sage
* remove model sparse
* upd
* upd
* remove global norm
* revert delete model_sparse.py
* update according to comments
* Fix doc
* upd
* Fix test
* upd
* lint
* lint
* lint
* upd
* upd
* clean up
Co-authored-by: Ubuntu <ubuntu@ip-172-31-56-220.ec2.internal >
2021-01-28 00:26:49 +08:00
a613ad8806
Refactor code for retaining formats in message-passing. ( #2570 )
2021-01-25 16:55:16 +08:00
25ac334403
[Distributed] Heterogeneous graph support ( #2457 )
...
* Distributed heterograph (#3 )
* heterogeneous graph partition.
* fix graph partition book for heterograph.
* load heterograph partitions.
* update DistGraphServer to support heterograph.
* make DistGraph runnable for heterograph.
* partition a graph and store parts with homogeneous graph structure.
* update DistGraph server&client to use homogeneous graph.
* shuffle node Ids based on node types.
* load mag in heterograph.
* fix per-node-type mapping.
* balance node types.
* fix for homogeneous graph
* store etype for now.
* fix data name.
* fix a bug in example.
* add profiler in rgcn.
* heterogeneous RGCN.
* map homogeneous node ids to hetero node ids.
* fix graph partition book.
* fix DistGraph.
* shuffle eids.
* verify eids and their mappings when loading a partition.
* Id map from homogneous Ids to per-type Ids.
* verify partitioned results.
* add test for distributed sampler.
* add mapping from per-type Ids to homogeneous Ids.
* update example.
* fix DistGraph.
* Revert "add profiler in rgcn."
This reverts commit 36daaed8b660933dac8f61a39faec3da2467d676.
* add tests for homogeneous graphs.
* fix a bug.
* fix test.
* fix for one partition.
* fix for standalone training and evaluation.
* small fix.
* fix two bugs.
* initialize projection matrix.
* small fix on RGCN.
* Fix rgcn performance (#17 )
Co-authored-by: Ubuntu <ubuntu@ip-172-31-62-171.ec2.internal >
* fix lint.
* fix lint.
* fix lint.
* fix lint.
* fix lint.
* fix lint.
* fix.
* fix test.
* fix lint.
* test partitions.
* remove redundant test for partitioning.
* remove commented code.
* fix partition.
* fix tests.
* fix RGCN.
* fix test.
* fix test.
* fix test.
* fix.
* fix a bug.
* update dmlc-core.
* fix.
* fix rgcn.
* update readme.
* add comments.
Co-authored-by: Ubuntu <ubuntu@ip-172-31-2-202.us-west-1.compute.internal >
Co-authored-by: Ubuntu <ubuntu@ip-172-31-9-132.us-west-1.compute.internal >
Co-authored-by: xiang song(charlie.song) <classicxsong@gmail.com >
Co-authored-by: Ubuntu <ubuntu@ip-172-31-62-171.ec2.internal >
* fix.
* fix.
* add div_int.
* fix.
* fix.
* fix lint.
* fix.
* fix.
* fix.
* adjust.
* move code.
* handle heterograph.
* return pytorch tensor in GPB.
* remove some tests in example.
* add to_block for distributed training.
* use distributed to_block.
* remove unnecessary function in DistGraph.
* remove distributed to_block.
* use pytorch tensor.
* fix a bug in ntypes and etypes.
* enable norm.
* make the data loader compatible with the old format.
* fix.
* add comments.
* fix a bug.
* add test for heterograph.
* support partition without reshuffle.
* add test.
* support partition without reshuffle.
* fix.
* add test.
* fix bugs.
* fix lint.
* fix dataset.
* fix for mxnet.
* update docstring.
* rename to floor_div
* avoid exposing NodePartitionPolicy and EdgePartitionPolicy.
* fix docstring.
* fix error.
* fixes.
* fix comments.
* rename.
* rename.
* explain IdMap.
* fix docstring.
* fix docstring.
* update docstring.
* remove the code of returning heterograph.
* remove argument.
* fix example.
* make GraphPartitionBook an abstract class.
* fix.
* fix.
* fix a bug.
* fix a bug in example
* fix a bug
* reverse heterograph sampling.
* temp fix.
* fix lint.
* Revert "temp fix."
This reverts commit c450717b9f578b8c48769c675f2a19d6c1e64381.
* compute norm.
* Revert "reverse heterograph sampling."
This reverts commit bd6deb7f52998de76508f800441ff518e2fadcb9.
* fix.
* move id_map.py
* remove check
* add more comments.
* update docstring.
Co-authored-by: Ubuntu <ubuntu@ip-172-31-2-202.us-west-1.compute.internal >
Co-authored-by: Ubuntu <ubuntu@ip-172-31-9-132.us-west-1.compute.internal >
Co-authored-by: xiang song(charlie.song) <classicxsong@gmail.com >
Co-authored-by: Ubuntu <ubuntu@ip-172-31-62-171.ec2.internal >
2021-01-24 19:18:00 -08:00
aa884d43b6
[doc][fix] Improve the docstring and fix its behavior in DGL's kernel ( #2563 )
...
* upd
* fix
* lint
* fix
* upd
2021-01-25 00:18:57 +08:00
917a755291
[Fix] fix version check ( #2544 )
2021-01-19 19:35:43 +08:00
f960468f50
[feature] APIs for manually set batch_num_nodes and batch_num_edges ( #2430 )
...
* wip
* udp
* upd
* fix typo
* lint
* lint
* upd
* upd
* lint
* lint
* upd
* upd
* tftest
* fix
* fallback numpy
* fix tf
* docstring
Co-authored-by: Mufei Li <mufeili1996@gmail.com >
2021-01-04 23:43:58 +08:00
e805470101
addressing post-merge comments ( #2455 )
2020-12-26 17:58:15 +08:00
9a7235faf2
[Performance] Use allocator from PyTorch if possible ( #2328 )
...
* first commit
* some thoughts
* move around
* more commit
* more fixes
* now it uses torch allocator
* fix symbol export error
* fix
* fixes
* test fix
* add script
* building separate library per version
* fix for vs2019
* more fixes
* fix on windows build
* update jenkinsfile
* auto copy built dlls for windows
* lint and installation guide update
* fix
* specify conda environment
* set environment for ci
* fix
* fix
* fix
* fix again
* revert
* fix cmake
* fix
* switch to using python interpreter path
* remove scripts
* debug
* oops sorry
* Update index.rst
* Update index.rst
* copies automatically, no need for this
* do not print message if library not found
* tiny fixes
* debug on nightly
* replace add_compile_definitions to make CMake 3.5 happy
* fix linking to wrong lib for multiple pytorch envs
* changed building strategy
* fix nightly
* fix windows
* fix windows again
* setup bugfix
* address comments
* change README
2020-12-25 13:57:51 +08:00
492ad9be81
[Fix] Avoid Overloading Nonzero for PyTorch Backend ( #2434 )
...
* Update gatedgraphconv.py
* Update entity_classify.py
* Update data-process.rst
* Update reading_data.py
* Update data-process.rst
* Update utils.py
* Update knowledge_graph.py
* Update entity_classify.py
* Update rdf.py
* Update entity_classify_mb.py
* Update test_classify.py
* Update tensor.py
* Update sparse.py
* Update entity_classify_mp.py
* Update 6_line_graph.py
2020-12-20 13:12:17 +08:00
6b02babbad
[doc] Add docstring for segment reduce. ( #2375 )
2020-11-27 18:04:13 +08:00
58775ada1f
Remove redundant print information in #2361 ( #2362 )
2020-11-23 03:41:58 +08:00
3adbfa1812
[Performance] Use segment operators for graph readout. ( #2361 )
...
* upd
* upd
* update
* upd
* upd
* upd
* fix
* lint
* lint
* pylint
* doc
2020-11-23 01:45:33 +08:00
5f19876358
[Bugfix] GINDataset check whether labels are all valid ( #2319 )
...
* PPIDataset
* Revert "PPIDataset"
This reverts commit 264bd0c960cfa698a7bb946dad132bf52c2d0c8a.
* check whether the original labels are valid
* add unit test for gin dataset
* bug of asscalar in mxnet<=1.6
* mxnet<=1.6 asscalar requires ndarray to be shaped (1,)
* skip gpu while testing datasets
* Update test_data.py
* test of gin dataset takes too much time
2020-11-05 09:35:10 +08:00
c6890c231a
[Performance] In HeteroNodeView, build arange on target device, instead of on CPU and copying it ( #2266 )
...
* Build arange on target device
* Utilize arange device in viewpy:HeteroNodeView.__call__
* Work around uint64 error in TF to_dlpack
* Restore else clause
Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com >
Co-authored-by: Da Zheng <zhengda1936@gmail.com >
2020-11-02 14:03:26 -08:00
c341520d5b
[bugfix] Fix the behavior of min/max reducer for 1-dim dta. ( #2250 )
...
* udp
* add test
* udp
* fix mxnet
2020-09-29 17:27:16 +08:00
7d8522a29f
[Bug] fix cumsum on an empty array with prepend_zero returning an empty array ( #2179 )
...
* fix cumsum
* udp
Co-authored-by: Zihao <expye@outlook.com >
2020-09-11 18:23:44 +08:00
cbd55eb12f
[performance] Batch DGLGraph in C++ end. ( #2155 )
...
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* fix
* upd
* upd
* upd
* upd
* fix
* upd
Co-authored-by: VoVAllen <jz1749@nyu.edu >
2020-09-11 02:54:27 +08:00
45b610c49a
fix edge_softmax ( #2160 )
...
Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com >
2020-09-09 14:19:43 +08:00
2ce426d956
[Bug] add_edges() crashes if the input tensor is empty ( #2100 )
...
* [Bug] add_edges() crashes if the input tensor is empty
* lint
* fix
2020-08-24 16:04:41 +08:00
35e25914ad
Update version check of PyTorch to 1.5 ( #2079 )
2020-08-20 21:33:29 +08:00
63e2ba23c6
[WIP][Kernel] Set the built-in reduce result of zero-degree nodes to 0 in C ( #2017 )
...
* test idea
* cuda kernels
* lint and fixes
* lint
* change to another strategy
* use infinity
* fix
Co-authored-by: Zihao Ye <expye@outlook.com >
2020-08-14 22:59:24 +08:00
6294677f8a
[hotfix] Set reduce results to all zero for nodes with zero in-degrees. ( #2011 )
2020-08-13 01:20:56 +08:00
f5eb80d221
[Feature] Edge DataLoader for edge classification & link prediction ( #1828 )
...
* clean commit
* oops forgot the most important files
* use einsum
* copy feature from frontier to block
* Revert "copy feature from frontier to block"
This reverts commit 5224ec963eb6a3ef1b6ab74d8ecbd44e4e42f285.
* temp fix
* unit test
* fix
* revert jtnn
* lint
* fix win64
* docstring fixes and doc indexing
* revert einsum in sparse bidecoder
* fix some examples
* lint
* fix due to some tediousness in remove_edges
* addresses comments
* fix
* more jtnn fixes
* fix
2020-08-11 18:00:58 +08:00
18bfec24cf
[hotfix] Refactor edge softmax module ( #1967 )
...
* upd
* upd
* upd
* upd
* upd
* upd
* upd
* upd
2020-08-07 23:50:12 +08:00
22167f7203
[Refactor] Enable new kernel in all message passing APIs ( #1953 )
...
* WIP: frame refactor
* new frame
* simple update_all builtin
* move all subgraph routines into the same file
* sddmm & spmm schedule; node & edge udf
* degree bucketing
* some tricky 0deg corner cases
* bug in frame append
* merge test_hetero_basics and test_basics
* some code rearange
* fix test_heterograph
* add mean spmm
* enable all builtin combinations
* pass gpu test
* pass pytorch tests
* wip
* fix some pt debugging codes
* fix bug in mxnet backward
* pass all mxnet utests
* passed tf tests
* docstring
* lint
* lint
* fix broadcasting bugs
* add warning and clamp for mean reducer
* add test for zero-degree mean
* address comments
* lint
* small fix
2020-08-07 15:40:25 +08:00
f9bde91fbd
[hotfix] Remove dead code ( #1942 )
...
* Remove dead code
* Fix
Co-authored-by: Ubuntu <ubuntu@ip-172-31-51-214.ec2.internal >
2020-08-05 23:16:46 +08:00
c801a164ee
[hotfix] Allow broadcastable expand in the backward phase of gspmm ( #1939 )
...
* upd
* upd
* upd
* upd
* upd
* trigger
* simplify unittest
* patch-sp
Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com >
2020-08-05 17:02:25 +08:00
a28bfa9f5f
[hotfix] Activate kernel unittest for tensorflow ( #1895 )
...
* upd
* upd
* upd
* upd
* upd
* trigger
* simplify unittest
Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com >
2020-07-31 19:57:53 +08:00
444becf00a
[Misc] Move many deprecated codes to the deprecate folder ( #1893 )
...
* some file movements
* move some codes to deprecated
* more deprecation
* lint
* remove useless test
2020-07-30 18:17:33 +08:00
Quan (Andy) Gan