TensorMap: Add ListKeys Op to TensorMap #2
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
Updates LLVM usage to match [e75bc5c791e0](llvm/llvm-project@e75bc5c791e0) PiperOrigin-RevId: 327449456 Change-Id: I847530d2d7325bd7bfeef29a3899d8cbcb3257e0
PiperOrigin-RevId: 327450338 Change-Id: I85b80f0254a7e0de0e2c206ee860d63358a3e5c4
…he handle. PiperOrigin-RevId: 327453287 Change-Id: I2b41af223f722920fcd1f9a0e5e1779f603412c0
…on/tf.WhileRegion and vice versa. Not preserving the device may result in placer using a different device than originally intended for such control flow ops. PiperOrigin-RevId: 327454861 Change-Id: I9e47be5fb21b36afe27965d1d0214c3df65c2231
PiperOrigin-RevId: 327455603 Change-Id: Ic5efe1f80fc7a92debbc4f08853f824f8cdfb937
PiperOrigin-RevId: 327456357 Change-Id: I32419edbc29b9b6a1f94f3fdcf91c053b7680c6d
…ward-compatibility with the support of multi-replica per sync-group. PiperOrigin-RevId: 327458432 Change-Id: Ie4d096558a58f81a7ea81f459294a3b10702bbb0
PiperOrigin-RevId: 327459673 Change-Id: I37c6f5dba9875c77eb8b784216a80d904472e5c0
PiperOrigin-RevId: 327461023 Change-Id: Ic61108261fa2084f7f81b8b52b7658e5f64c0a0b
PiperOrigin-RevId: 327462578 Change-Id: I40b493972dade171145dd2b245a270bf1ecbfdfe
PiperOrigin-RevId: 327463012 Change-Id: Ib8d9eaaa4f5897b642bc770828e7e2d09b76d7fb
- ToBool is always expected to produce a single scalar value. Constrain the ODS definition to enforce this constraint PiperOrigin-RevId: 327463327 Change-Id: I7769b7f3191c74998c6098a3f2d41660394ca9e0
…ta C++ classes. PiperOrigin-RevId: 327468816 Change-Id: Ib030b0d2636075cacd084014552b317b0654cd5c
PiperOrigin-RevId: 327470035 Change-Id: Iee0e7311aa76ec44658f458ea59e23cfdcddbd51
PiperOrigin-RevId: 327470626 Change-Id: I36de07a757b3d4fda8c3d53f453bead21f6c8535
…s in XLA that can result in different accumulation orders for reduction operations. As a result, relax the precision of the keras correctness tests. PiperOrigin-RevId: 327474361 Change-Id: I8fcf1a2889d6c99f5c757fc8f6a57d1f20cb973f
…t classes. Rename only; no functional changes. PiperOrigin-RevId: 327476904 Change-Id: I85a410bd68ad9192d5a43107b94cad4e1aeb83f0
No functionality change. PiperOrigin-RevId: 327477678 Change-Id: I2d423afba5f18ef28b89b75c22db1c75cad580f3
PiperOrigin-RevId: 327479427 Change-Id: I325b5f29e32fbaeb44de0013da32bce1e41fcecb
…ialize multi-process runner so it can be caught if desired. PiperOrigin-RevId: 327480135 Change-Id: Ic8e8ecb432fe8d216c373a2c69d5e94417d67723
Fixed typos
PiperOrigin-RevId: 327482183 Change-Id: I1e91e50905cb7011fe987a40a65688f2ef1d091c
PiperOrigin-RevId: 327483991 Change-Id: I1d934db2e61ebbd7aa4fe0010fd52d4658e8a09d
Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
PiperOrigin-RevId: 327492744 Change-Id: Ie899823ea66e68e15fbda3578acd9cf5893554cb
…ToDeviceMetadata. PiperOrigin-RevId: 327492950 Change-Id: I0b8fcd3ff46e683639d99db0e12ad9e94d6b7414
This is now disabled by default but can let us test the idea. The thread keeps checking the cluster and aborts the collective if any worker is not reachable. Currently the only way to recovery is to restart, since once collectives are aborted, all subsequent collectives fail immediately. The thread uses a RING all-reduce as a check mechanism, since we don't have a check alive op yet. PiperOrigin-RevId: 327493026 Change-Id: I9ab9e5be1f5c1a15b3741a9f26e42f25b9a59a12
PiperOrigin-RevId: 327494182 Change-Id: I70a8d649b51c891d720b789db75befed359ec08f
PiperOrigin-RevId: 327999616 Change-Id: Ie46cc103cad75561bd863c8d477eaa5c9f319452
PiperOrigin-RevId: 328002300 Change-Id: I5713bc15bb0d7a8647b1097fe81570ace30cb1c5
PiperOrigin-RevId: 328009562 Change-Id: I462f5a36c28b5e71ce562e67c3226feea5d2fb7d
PiperOrigin-RevId: 328009565 Change-Id: If7bb6f781e9a5381e35e98408272a5cdb42f64c8
MLIR is moving to require explicitly loading of Dialect before creating entities in a Dialect. PiperOrigin-RevId: 328037037 Change-Id: Ib46275b26e8f77aab0fbd0f70cd2a48844dc360c
Imported from GitHub PR tensorflow#42508 An lmhlo.constant op on an memref that is locally allocated and with no users other than dealloc's can be deleted. Add a canonicalization pattern for this. Copybara import of the project: -- 8758c40 by Uday Bondhugula <uday@polymagelabs.com>: [MLIR] Erase dead lmhlo.constant ops An lmhlo.constant op on an memref that is locally allocated and with no other users (other than dealloc's) can be deleted. Add a canonicalization patter for this. COPYBARA_INTEGRATE_REVIEW=tensorflow#42508 from polymage-labs:lhlo_constant_erase 8758c40 PiperOrigin-RevId: 328042416 Change-Id: I27f9b5b5297bbf6fe81aff589f009197b75f49eb
1. Try to reuse the original target tiled sharding when finding compatible target from partial sharding. 2. If the HLO is a broadcast, check if data is already the same between source/target pairs. PiperOrigin-RevId: 328043490 Change-Id: I69dec53c50cb6cedf586afafc5181cd1ee29cdc6
Updates LLVM usage to match [f164534ca8e0](llvm/llvm-project@f164534ca8e0) PiperOrigin-RevId: 328046788 Change-Id: I714164211a50e0d273ec49046c66f7e484989428
PiperOrigin-RevId: 328068227 Change-Id: Ia084d946f3a0e5d071d7e8fec4263d1da26d9671
…lver for unit tests of xnnpack delegate itself to prepare for enabling xnnpack delegate by default across all platforms in the next 2.4.0 release. PiperOrigin-RevId: 328068258 Change-Id: I3459bc3e7f25d2925da65fba3e19ac2bad57fff1
Updates LLVM usage to match [f6decfa36d89](llvm/llvm-project@f6decfa36d89) PiperOrigin-RevId: 328073633 Change-Id: I5cd74bcf36c453cf073766f910a0f8442b66cb93
…lver for unit tests of xnnpack delegate itself to prepare for enabling xnnpack delegate by default across all platforms in the next 2.4.0 release. PiperOrigin-RevId: 328076934 Change-Id: I69e21a6fbbe1b0e7146669ccd6481b774dcd9d2e
TFE_TensorHandleToNumpy seems to be working with TFRT. PiperOrigin-RevId: 328077896 Change-Id: I0fc569e67440c00327009d87e50960734ed2bce1
Updates LLVM usage to match [b999400a4fb6](llvm/llvm-project@b999400a4fb6) PiperOrigin-RevId: 328096769 Change-Id: I42c48819078ea634ec7a7bf4d73579c7ddda9c47
PiperOrigin-RevId: 328100698 Change-Id: Ia54de53b1a7abf269de44d5aa1817c14b9da1053
PiperOrigin-RevId: 328100708 Change-Id: Ib79b4c72b07ce8c03bdaaac246f4bc492a116914
…mpiler/mlir/lite/... PiperOrigin-RevId: 328109152 Change-Id: Ia460e89f785e9a2aaf21538083733e7e13730299
…t one consumer. PiperOrigin-RevId: 328128307 Change-Id: Ia9a5e9f4c20d78deee32b738ae1002d80eb935c1
PiperOrigin-RevId: 328129947 Change-Id: If3fc9bd3ccb4a271c60e48e052969c07174326e4
PiperOrigin-RevId: 328133653 Change-Id: I99ceafa58d99364665175222748ed1a4ae16b0e4
PiperOrigin-RevId: 328134070 Change-Id: I3ad10cd9ee442e7f478192c3ea0e1254077c5f2e
…nt8_float PiperOrigin-RevId: 328137876 Change-Id: I809b345ca9c54c352959987c904d122347436aa6
…-var PiperOrigin-RevId: 328140122 Change-Id: Iacbb71fb3288057de3db6e0ab06e1cea20961223
…ta C++ classes. PiperOrigin-RevId: 328141067 Change-Id: I154c55c9b1cc47ec069bfc663a1d3fb0f135e67d
Updates LLVM usage to match [bad7d6b3735d](llvm/llvm-project@bad7d6b3735d) PiperOrigin-RevId: 328142991 Change-Id: I54a09bd386149cd5a649f5e0ca7bcc97028692c5
…::ConstOp PiperOrigin-RevId: 328149666 Change-Id: I0c5561152383f12126ab9568c0facc4c3043c6a3
PiperOrigin-RevId: 328150132 Change-Id: If1468b900e95398106fcca29d0263ebb8869731e
…ssfully. PiperOrigin-RevId: 328157361 Change-Id: I57675a6822cdd56c662aa877077d4688aa6a4411
Remove const type from vectors. PiperOrigin-RevId: 328157979 Change-Id: I8df58b0b23831b842c04c3243290ca61ecf7f4aa
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.
Add another op for the TensorMap feature: TensorMapListKeys. Given a TensorMap handle and key dtype, TensorMapListKeys returns a Tensor containing all keys in the map. Code changes include adding and registering op kernel, wrapping op in python, and testing.