Branch 23.06 Fix Forward Merge #3462
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
A CAPI implementation of betweenness centrality is available and this PR: 1. Implement PLC betweenness centrality by leveraging the CAPI 2. Refactor the python SG implementation of betweenness centrality by leveraging the PLC implementation 3. Add a python MG implementation of betweenness centrality with tests closes rapidsai#3145 closes rapidsai#2605 closes rapidsai#2648 closes rapidsai#2649 closes rapidsai#2650 Authors: - Joseph Nke (https://github.com/jnke2016) - Chuck Hastings (https://github.com/ChuckHastings) - Rick Ratzel (https://github.com/rlratzel) - Alex Barghi (https://github.com/alexbarghi-nv) - Vyas Ramasubramani (https://github.com/vyasr) - Vibhu Jawa (https://github.com/VibhuJawa) Approvers: - Rick Ratzel (https://github.com/rlratzel) - Chuck Hastings (https://github.com/ChuckHastings) - Alex Barghi (https://github.com/alexbarghi-nv) - Brad Rees (https://github.com/BradReesWork) URL: rapidsai#2971
This PR adds a working Multi-GPU Graph (on 2 dask workers) being trained/loaded on multiple pytorch trainers. (3) Todo: - [x] Verify works on multiple trainers and multiple dask workers - [x] Show scaling as you increase training GPUs At 1 second we become bottlenecked by sampling dask cluster, but we see perf improvement by going from `1 GPU`->`2GPU`. **On OBGN-Products** ```md | Number of Training GPUs | Time per epoch | |-------------------------|----------------| | 1 | 2.3 s | | 2 | 0.582 s | | 4 | 0.792 s | ``` This PR depends upon: rapidsai#3393 CC: @rlratzel , @alexbarghi-nv , @BradReesWork Authors: - Vibhu Jawa (https://github.com/VibhuJawa) - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - Alex Barghi (https://github.com/alexbarghi-nv) URL: rapidsai#3212
resolves rapidsai#3297 new website structure with some migrations. Authors: - Don Acosta (https://github.com/acostadon) - Brad Rees (https://github.com/BradReesWork) Approvers: - Brad Rees (https://github.com/BradReesWork) - Alex Barghi (https://github.com/alexbarghi-nv) URL: rapidsai#3343
…dsai#3289) Updates tests to support presence of PyTorch and PyG in the environment. Updates cugraph-pyg to include support for loading from samples on disk. Makes CuGraphStore serializable and allows creating CuGraphStore and BulkSampleLoader instances without the graph structure. Merge after rapidsai#3288 Closes rapidsai#3287 Closes rapidsai#3176 Authors: - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - Vibhu Jawa (https://github.com/VibhuJawa) - Rick Ratzel (https://github.com/rlratzel) URL: rapidsai#3289
This PR uses dependencies.yaml to generate the dependency lists in pyproject.toml Authors: - Vyas Ramasubramani (https://github.com/vyasr) Approvers: - AJ Schmidt (https://github.com/ajschmidt8) - Brad Rees (https://github.com/BradReesWork) URL: rapidsai#3355
The `__version__` attribute appears to have been accidentally deleted in rapidsai#2971. Authors: - Vyas Ramasubramani (https://github.com/vyasr) Approvers: - Alex Barghi (https://github.com/alexbarghi-nv) URL: rapidsai#3411
Authors: - Artur (https://github.com/ArturKasymov) - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - Brad Rees (https://github.com/BradReesWork) URL: rapidsai#3308
…cugraph-pyg (rapidsai#3382) Resolves an issue where the wrong version of `searchsorted` caused a device to host copy. Also removes the backend option from `CuGraphStore` entirely to prevent similar bugs from happening in the future and better align cugraph-pyg with the pyg/pytorch ecosystem. Merge after rapidsai#3289 Closes rapidsai#2995 Authors: - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - Rick Ratzel (https://github.com/rlratzel) - Vibhu Jawa (https://github.com/VibhuJawa) URL: rapidsai#3382
…#3393) This PR fixes a bug where output sample batch ids do not match those expected when using the bulk sampler, causing subgraphs that are larger than expected and incorrect. Without reindexing, the wrong batch ids are assigned to the start vertices. Reindexing ensures that the same order is preserved for batch ids and start vertices. This PR also changes the empty dataframe passed to dask in `uniform_neighbor_sample` to match the correct ordering of batch_id and hop_id. This ensures that the columns are named correctly and are not inadvertently renamed due to them being created in a different order. This PR is non-breaking because it restores the original behavior of bulk sampling and reverses a bug that was inadvertently introduced with the dask updates. Resolves rapidsai#3390 Authors: - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - Rick Ratzel (https://github.com/rlratzel) - Vibhu Jawa (https://github.com/VibhuJawa) - Joseph Nke (https://github.com/jnke2016) URL: rapidsai#3393
This PR adds an MG implementation of induced subgraph by leveraging the CAPI closes rapidsai#2535 closes rapidsai#2536 Authors: - Joseph Nke (https://github.com/jnke2016) - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - Chuck Hastings (https://github.com/ChuckHastings) - Rick Ratzel (https://github.com/rlratzel) URL: rapidsai#3391
### Leiden algorithm This PR adds Leiden implementation using cugraph primitives. It also add an implementation of `maximal independent set ` using cugraph primitives that is used by the Leiden implementation. - It reuses code from Louvain implementation to find locally optimal moves for vertices. - Locally optimal movement creates clusters of `Louvain communities` - In refinement phase, which is the main feature of Leiden algorithm, vertices are moved only within a `Louvain community` to subdivide a `Louvain community` in multiple `sub communities` or `Leiden communities` - In graph contraction phase, vertices belonging to a `Leiden community` are merged to become a node in aggregated graph. This code has been tested to work on SG. We aim to test MG version in next release. Future improvement: - Move vertices randomly instead of greedily refinement phase - Test for MNMG Authors: - Naim (https://github.com/naimnv) - Chuck Hastings (https://github.com/ChuckHastings) Approvers: - Chuck Hastings (https://github.com/ChuckHastings) - Seunghwa Kang (https://github.com/seunghwak) URL: rapidsai#2980
- Adds new examples for cugraph-pyg. - Removes outdated examples. - Moves MG scripts to top-level directory. - Makes the input to `_get_vertex_groups_from_sample` a tensor instead of Series - Adds `is_sorted` arg to `_get_vertex_groups_from_sample` to skip sorting if tensor already sorted - Some fixes to `CuGraphStore` for running multi-GPU workflows Merge after rapidsai#3288 - merged Merge after rapidsai#3289 - merged Merge after rapidsai#3382 - merged Closes rapidsai#3316 Closes rapidsai#3226 Closes rapidsai#3072 Authors: - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - Vibhu Jawa (https://github.com/VibhuJawa) - Rick Ratzel (https://github.com/rlratzel) URL: rapidsai#3326
…3300) The allocator callbacks now live in their own submodules (so that RMM does not, for example, import pytorch unless required) and so must be explicitly imported. Authors: - Lawrence Mitchell (https://github.com/wence-) Approvers: - Alex Barghi (https://github.com/alexbarghi-nv) - Rick Ratzel (https://github.com/rlratzel) URL: rapidsai#3300
closes rapidsai#2437 closes rapidsai#1679 closes rapidsai#2535 closes rapidsai#2536 closes rapidsai#2537 Authors: - Joseph Nke (https://github.com/jnke2016) - Chuck Hastings (https://github.com/ChuckHastings) - Alex Barghi (https://github.com/alexbarghi-nv) - Rick Ratzel (https://github.com/rlratzel) Approvers: - Brad Rees (https://github.com/BradReesWork) - Chuck Hastings (https://github.com/ChuckHastings) - Alex Barghi (https://github.com/alexbarghi-nv) - Vibhu Jawa (https://github.com/VibhuJawa) - Rick Ratzel (https://github.com/rlratzel) - Ray Douglass (https://github.com/raydouglass) URL: rapidsai#2949
…rm Neighbor Sample (rapidsai#3416) Currently, cudf does not merge series properly when they already share an index. I'm not sure if this is a bug in cudf, or intentional behavior. This issue does not occur with dask_cudf. The resolution is to use `cudf.concat` when passing a `cudf.Series` for start vertices and batch ids, and `df.to_frame().merge` when passing in a `dask_cudf.Series` for start vertices and batch ids. This PR also adds an additional test which tests both cudf and dask_cudf inputs to catch these sort of problems in the future. Authors: - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - Rick Ratzel (https://github.com/rlratzel) - Vibhu Jawa (https://github.com/VibhuJawa) - Joseph Nke (https://github.com/jnke2016) URL: rapidsai#3416
…sai#3352) Fixes rapidsai#3058 Authors: - Erik Welch (https://github.com/eriknw) - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - Alex Barghi (https://github.com/alexbarghi-nv) URL: rapidsai#3352
Required to support the upcoming minor release of PyG, which will allow full compatibility with pylibcugraphops. Authors: - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - Vibhu Jawa (https://github.com/VibhuJawa) - Ray Douglass (https://github.com/raydouglass) URL: rapidsai#3422
Trying to fix doc build issues Authors: - Brad Rees (https://github.com/BradReesWork) Approvers: - Don Acosta (https://github.com/acostadon) - Rick Ratzel (https://github.com/rlratzel) URL: rapidsai#3418
…#3360) Update cugraph-ops models to use pylibcugraphops 23.04. This PR also supersedes rapidsai#3264 CC: @MatthiasKohl Authors: - Tingyu Wang (https://github.com/tingyu66) - Alex Barghi (https://github.com/alexbarghi-nv) Approvers: - AJ Schmidt (https://github.com/ajschmidt8) - Vibhu Jawa (https://github.com/VibhuJawa) - Rick Ratzel (https://github.com/rlratzel) URL: rapidsai#3360
This PR pins `dask` and `distributed` to `2023.3.2` and `2023.3.2.1` respectively for `23.04` release. xref: rapidsai/cudf#13070 Authors: - GALI PREM SAGAR (https://github.com/galipremsagar) Approvers: - Ray Douglass (https://github.com/raydouglass) - Peter Andreas Entschev (https://github.com/pentschev) - Rick Ratzel (https://github.com/rlratzel) - Joseph (https://github.com/jolorunyomi) URL: rapidsai#3427
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
alexbarghi-nv
added
bug
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Resolves the squash merge done accidentally.