Tags: andrewjrae/pytorch
Tags
[ZeRO] Add ctor support for multiple param groups (pytorch#72578) Summary: Pull Request resolved: pytorch#72578 **Overview** This adds `ZeroRedundancyOptimizer` constructor support for multiple parameter groups (i.e. passing an `iterable` of `dict`s instead of an `iterable` of `torch.Tensor` as the `parameters` argument) to mirror the API for non-sharded optimizers. Fixes pytorch#71347 and pytorch#59973. This modifies `test_collect_shards()` to skip if ROCm. **Test Plan** I adjusted the existing constructor test, and I added a test for parity between constructing with two parameter groups up front versus constructor with one parameter group and adding the second parameter group after (via `add_param_group()`) versus a non-sharded optimizer. Test Plan: Imported from OSS Reviewed By: rohan-varma Differential Revision: D34106940 Pulled By: awgu fbshipit-source-id: 7e70fc0b3cec891646e0698eaedf02ff4354c128 (cherry picked from commit 40f2d45)
Update on "[quant] Add ConvTranspose reference module - Reland pytorc… …h#73031" Summary: Add ConvTranspose reference module Test Plan: python3 test/test_quantization.py TestQuantizeEagerOps.test_conv_transpose_2d Reviewers: Subscribers: Tasks: Tags: Differential Revision: [D34352228](https://our.internmc.facebook.com/intern/diff/D34352228) [ghstack-poisoned]
Sparse CSR CPU: implement addmm(dense, sparse, sparse) -> dense This PR adds a possiblity to multiply two sparse matrices and add the result to a dense matrix. It uses an MKL function and only CPU path is implemented for now.
Enable CSR inputs for torch.sparse.mm Previously `torch.sparse.mm` supported only COO and dense inputs. Computing derivatives works wrt dense input for sparse_csr x dense -> dense Modified implementation of `torch.sparse.mm` to be directly bound to ATen function.
Merge remote-tracking branch 'upstream/viable/strict' into lu-solve-b… …atch-expand
Merge branch 'master' of https://github.com/mszhanyi/pytorch into zha… …nyi/ignoreioenv
Update on "Add BUILD_LAZY_CUDA_LINALG option" When enable, it will generate `torch_cuda_linalg` library, which would depend on cusolve and magma and registers dynamic bindings to it from LinearAlgebraStubs Differential Revision: [D33992795](https://our.internmc.facebook.com/intern/diff/D33992795) [ghstack-poisoned]
PreviousNext