 Ramon Zhou
|
f0213d2163
|
[GraphBolt] Refactor sampling (#7367)
|
2024-04-29 15:56:57 +08:00 |
|
|
Muhammed Fatih BALIN
|
6b140f28ed
|
[GraphBolt] Hetero CPU sampling bug fix. (#7369)
|
2024-04-29 00:29:41 -04:00 |
|
|
pyynb
|
0d9a09df18
|
[graphbolt] Disable liburing (#7366)
|
2024-04-29 11:48:06 +08:00 |
|
|
Ramon Zhou
|
658b2086b0
|
[GraphBolt] Optimize hetero sampling on CPU (#7360)
|
2024-04-28 13:51:27 +08:00 |
|
|
peizhou001
|
2251683443
|
[Temporal]Fix tensor option mismatch issue (#7346)
Co-authored-by: Ubuntu <ubuntu@ip-172-31-21-218.ap-northeast-1.compute.internal>
|
2024-04-23 16:28:42 +08:00 |
|
|
Muhammed Fatih BALIN
|
b68a2fc069
|
Revert "[GraphBolt] Quick fix for GCC 8" (#7328)
|
2024-04-19 01:40:28 -04:00 |
|
|
Muhammed Fatih BALIN
|
9a340c77b4
|
[GraphBolt][CUDA] Add missing thrust include for later CCCL versions. (#7325)
|
2024-04-19 10:58:15 +08:00 |
|
|
Muhammed Fatih BALIN
|
a3d20dce88
|
[GraphBolt][CUDA] Make _convert_to_sampled_subgraph lighter. (#7312)
|
2024-04-18 12:44:57 -04:00 |
|
|
Muhammed Fatih BALIN
|
4546e54f21
|
[GraphBolt][CUDA] Don't link against libcuda.so. (#7313)
|
2024-04-18 02:57:29 -04:00 |
|
|
Muhammed Fatih BALIN
|
f0f424f70c
|
[GraphBolt][CUDA] Better TORCH_CUDA_ARCH_LIST initialization. (#7293)
|
2024-04-10 13:41:53 +08:00 |
|
|
Muhammed Fatih BALIN
|
7b470eb0fa
|
[GraphBolt] Quick fix for GCC 8 (#7279)
|
2024-04-08 01:12:18 -04:00 |
|
|
Muhammed Fatih BALIN
|
78df81015a
|
[GraphBolt] Add optimized unique_and_compact_batched. (#7239)
|
2024-04-07 04:55:08 -04:00 |
|
|
Muhammed Fatih BALIN
|
7eb4de4bed
|
[GraphBolt][CUDA] Sample neighbors synchronization optimization. (#7264)
|
2024-04-07 02:38:51 -04:00 |
|
|
Muhammed Fatih BALIN
|
d4a6f8a036
|
[GraphBolt][CUDA] Refactor Gather operation. (#7269)
|
2024-04-06 23:45:58 +08:00 |
|
|
Muhammed Fatih BALIN
|
d3483fe1f1
|
[GraphBolt][CUDA] Optimize hetero sampling. (#7223)
|
2024-04-03 03:18:30 -04:00 |
|
|
xiangyuzhi
|
b725ee53d7
|
[Graphbolt] Disk feature fetcher (#7128)
Co-authored-by: Ubuntu <ubuntu@ip-172-31-40-250.ap-northeast-1.compute.internal>
|
2024-04-01 16:01:43 +08:00 |
|
|
Ramon Zhou
|
3c39153301
|
[GraphBolt] Make unique_and_compact deterministic (#7217)
|
2024-03-25 16:26:03 +08:00 |
|
|
Muhammed Fatih BALIN
|
9632ab1d7c
|
[GraphBolt][CUDA] Specialize non-weighted neighbor sampling impl (#7215)
|
2024-03-18 10:38:44 -04:00 |
|
|
Muhammed Fatih BALIN
|
a2c5472a74
|
[GraphBolt] Labor dependent template specialization. (#7220)
|
2024-03-18 03:08:31 -04:00 |
|
|
Muhammed Fatih BALIN
|
a272efed96
|
[GraphBolt] Implement labor dependent minibatching - python side. (#7208)
|
2024-03-13 13:05:19 -04:00 |
|
|
Muhammed Fatih BALIN
|
f0c7efa9a4
|
[GraphBolt] Implement dependent minibatching for labor. (#7205)
|
2024-03-11 08:23:36 -04:00 |
|
|
Muhammed Fatih BALIN
|
bc978736fc
|
[GraphBolt] torch.compile() support for gb.expand_indptr. (#7188)
|
2024-03-05 00:26:54 -05:00 |
|
|
Muhammed Fatih BALIN
|
8c28f98266
|
[GraphBolt] Slight gb.expand_indptr improvement on CPU. (#7189)
|
2024-03-03 23:08:41 -05:00 |
|
|
xiangyuzhi
|
0c3a4428c1
|
[Graphbolt] Add liburing dependency (#7149)
Co-authored-by: Ubuntu <ubuntu@ip-172-31-40-250.ap-northeast-1.compute.internal>
|
2024-02-29 14:48:55 +08:00 |
|
|
Muhammed Fatih BALIN
|
73e01d6daa
|
[CUDA] Update CCCL to 2.3.0 (#7171)
|
2024-02-29 00:37:00 -05:00 |
|
|
Mingbang Wang
|
f498986771
|
[GraphBolt] Add error messages for attribute size check (#7079)
Co-authored-by: Muhammed Fatih BALIN <m.f.balin@gmail.com>
|
2024-02-04 20:10:45 +08:00 |
|
|
czkkkkkk
|
1e6fa711f2
|
[Graphbolt] Add fast path for tamporal sampling. (#7078)
|
2024-02-03 17:05:58 +08:00 |
|
|
Muhammed Fatih BALIN
|
15695ed0ec
|
[GraphBolt][CUDA] Handle edge case of %100 cache hit rate. (#7080)
|
2024-02-02 23:03:06 -08:00 |
|
|
Muhammed Fatih BALIN
|
d5b03bcb27
|
[GraphBolt][CUDA] GPUCache performance fix. (#7073)
|
2024-02-02 14:49:54 -08:00 |
|
|
Muhammed Fatih BALIN
|
177dc133d6
|
[GraphBolt][CUDA] IndexSelectCSC kernel launch config change. (#7056)
|
2024-02-02 13:16:52 +08:00 |
|
|
Muhammed Fatih BALIN
|
3a79f02189
|
[GraphBolt][CUDA] Make nodes optional for sampling (#6993)
|
2024-01-24 12:58:54 -05:00 |
|
|
Muhammed Fatih BALIN
|
365bb723d9
|
[GraphBolt][CUDA] Optimize UVA index_select_csc. (#6995)
|
2024-01-24 12:30:55 -05:00 |
|
|
Muhammed Fatih BALIN
|
aad12df6a6
|
[GraphBolt][CUDA] gb.expand_indptr (#6871)
|
2024-01-18 04:11:01 -05:00 |
|
|
Muhammed Fatih BALIN
|
f6db850d75
|
[GraphBolt][CUDA] Add native GPUCachedFeature instead of using DGL (#6939)
|
2024-01-18 02:59:35 -05:00 |
|
|
Muhammed Fatih BALIN
|
528b041c51
|
[GraphBolt][CUDA] Overlap feature fetcher (#6954)
|
2024-01-18 02:58:54 -05:00 |
|
|
Muhammed Fatih BALIN
|
0f3bfd7ecb
|
[GraphBolt][CUDA] Refactor IndexSelectCSC and add output_size argument (#6927)
|
2024-01-12 06:10:26 -05:00 |
|
|
Muhammed Fatih BALIN
|
3795a006b9
|
[GraphBolt][CUDA] Refactor codebase with CUB_CALL macro (#6870)
|
2024-01-12 05:00:24 -05:00 |
|
|
Muhammed Fatih BALIN
|
f86212edb5
|
[GraphBolt][CUDA] Enable tests for weighted sampling (#6919)
|
2024-01-12 03:51:53 -05:00 |
|
|
peizhou001
|
afca111427
|
[Graphbolt]Fix negative sampler (#6933)
Co-authored-by: Ubuntu <ubuntu@ip-172-31-21-218.ap-northeast-1.compute.internal>
|
2024-01-11 16:29:50 +08:00 |
|
|
Muhammed Fatih BALIN
|
1193e2e8ec
|
[GraphBolt][CUDA] Dataloader num_workers > 0 fix. (#6924)
|
2024-01-10 11:27:05 +08:00 |
|
|
Muhammed Fatih BALIN
|
b3224ce8db
|
[GraphBolt] Switch to using AT_DISPATCH_INDEX_TYPES for graph (#6912)
|
2024-01-08 02:27:17 -05:00 |
|
|
Muhammed Fatih BALIN
|
61504ec5e6
|
[GraphBolt] Extend temporal sampling to labor (#6816)
|
2024-01-08 01:04:27 -05:00 |
|
|
Muhammed Fatih BALIN
|
557a8f817a
|
[GraphBolt][CUDA] Multiple fanout hetero sampling support. (#6895)
|
2024-01-05 01:58:15 -05:00 |
|
|
Muhammed Fatih BALIN
|
25eabbb4ef
|
[GraphBolt] Setting manual seed also sets the seed now. (#6900)
|
2024-01-04 22:26:26 -05:00 |
|
|
czkkkkkk
|
29c3b06d4a
|
[Graphbolt] Change the temporal filter condition of temporal sampler. (#6893)
|
2024-01-04 11:20:10 +08:00 |
|
|
Muhammed Fatih BALIN
|
ce223b3693
|
[GraphBolt][CUDA] Add type_per_edge reading support and ensure same types are consecutive (#6865)
Co-authored-by: Rhett Ying <85214957+Rhett-Ying@users.noreply.github.com>
|
2024-01-01 21:53:28 -05:00 |
|
|
peizhou001
|
a2cb2ecd0e
|
[Graphbolt]Fix windows nominmax issue (#6853)
Co-authored-by: Ubuntu <ubuntu@ip-172-31-21-218.ap-northeast-1.compute.internal>
|
2023-12-29 09:48:57 +08:00 |
|
|
Muhammed Fatih BALIN
|
22a2513d0a
|
[GraphBolt][CUDA] Reduce and hide unique_and_compact synchronizations. (#6841)
|
2023-12-28 11:16:21 -05:00 |
|
|
Muhammed Fatih BALIN
|
e7f0c3a160
|
[GraphBolt][CUDA] SampleNeighbors (Without replacement for now) (#6770)
|
2023-12-28 10:59:20 -05:00 |
|
|
Muhammed Fatih BALIN
|
01df9bade9
|
[GraphBolt][CUDA] InSubgraph (#6830)
|
2023-12-28 09:52:28 -05:00 |
|