- 26 6月, 2020 19 次提交
-
-
由 Austin Anderson 提交于
PiperOrigin-RevId: 318325499 Change-Id: I32ca160bd0ecc01676b8758f3496415df8b7268c
-
由 Shanqing Cai 提交于
PiperOrigin-RevId: 318320055 Change-Id: I26271c9339b0381dd624a8dcee8758083b1313a8
-
由 Marcello Maggioni 提交于
PiperOrigin-RevId: 318314472 Change-Id: I9cc5a2e7ad9fefe85ef0099d2b9d33c31b5f405d
-
由 TensorFlower Gardener 提交于
PiperOrigin-RevId: 318313919 Change-Id: Idc03617514f781477874f15242d8fe9f3375d0f8
-
由 Ruoxin Sang 提交于
Fix a memory leak in TPUStrategy. Currently TPUStrategy always caches the function passed into `strategy.run`, which will causes objects not released in time. PiperOrigin-RevId: 318312817 Change-Id: I7b275fe87a4454d94ce31076ce38d12a635c477b
-
由 Katherine Wu 提交于
OP Layers wrap a single Tensorflow op in a Layer class. Previously, SavedModel would wrap every internal layer call in a tf.function, so that the user can inspect individual layers in the loaded model. For TensorflowOpLayer, this is unnecessary because (1) wrapping a single op in a tf.function is very inefficient (2) the user is unlikely to individually inspect the autogenerated op layers in the loaded model. This change also resolves the saving issue that occurs when a user builds a functional model while using the eager-computed results of `tf.shape(x)` as the input shape to another op layer. An example to help illustrate: ``` x = tf.keras.Input((2,)) # Shape is (None, 2) state = tf.zeros(4, tf.shape(x)[0]) # Expected shape is (4, None) LSTM(inputs, initial_state=state) ``` Prior to this CL, the TensorFlowOpLayers generated for tf.shape and tf.zeros would be separately wrapped in tf.functions when saving. This results in `state` having a shape of `(None, None)` instead of `(4, None)`, causing potential problems when saving the rest of the model. PiperOrigin-RevId: 318311978 Change-Id: I15099d8ba29c1d4facd3f88630f8e2651f22ae83
-
由 Andy Ly 提交于
PiperOrigin-RevId: 318307554 Change-Id: I0fdaab87f8fa98696e02269d4bd9b12c6e8a3fa4
-
由 TensorFlower Gardener 提交于
PiperOrigin-RevId: 318305127 Change-Id: I7d44bc8a273cbb27b872bf05480f3bf9676289a8
-
由 TensorFlower Gardener 提交于
PiperOrigin-RevId: 318304520 Change-Id: I33d49dfbcf9ffb080a4cb3a199563f43c0c948ec
-
由 Giorgio Arena 提交于
Imported from GitHub PR https://github.com/tensorflow/tensorflow/pull/34999 …ion != 1 Copybara import of the project: -- e6eabb04686999aebae95d2964d5f3dc3ac96061 by Giorgio Arena <giorgio.arena@arm.com>: Extend tests in TFLite micro for conv and depth_conv to support dilation != 1 -- 14cee9716ca0622b96aca6f6d84b9320420954bc by Giorgio Arena <giorgio.arena@arm.com>: Fix depthwise_conv dilated mismatches COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/tensorflow/pull/34999 from giorgio-arenarm:dilation_tests_tflu 14cee9716ca0622b96aca6f6d84b9320420954bc PiperOrigin-RevId: 318304398 Change-Id: I04e16c34fdcc123e6d93dc15ff021fa69eadda6e
-
由 Frank Chen 提交于
PiperOrigin-RevId: 318303930 Change-Id: I2b3ca2b2d3b32276bec73cc35c3d550ab596b4d8
-
由 Marcello Maggioni 提交于
PiperOrigin-RevId: 318302332 Change-Id: I25d3be03fbcca34d62e4f8d6b150926974e26f95
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 318301423 Change-Id: I60a6616077a2e27b1bc06b66d37af19fab11673d
-
由 Raman Sarokin 提交于
PiperOrigin-RevId: 318296675 Change-Id: If29361fb81f29b08fdcc094cc0ad9b60a97a4d67
-
由 Lei Zhang 提交于
PiperOrigin-RevId: 318294378 Change-Id: I3bfb10e133133b3651015f1a71078fc5e1d926b3
-
由 TensorFlower Gardener 提交于
PiperOrigin-RevId: 318293895 Change-Id: I5d65f9cc3b764d0136fa74992445a53b83dfbe08
-
由 T.J. Alumbaugh 提交于
PiperOrigin-RevId: 318293374 Change-Id: Iab2333401127d362c136940cd54877f508a0d6cd
-
由 Yanhua Sun 提交于
The same test can run with and without XLA compilation. In non-XLA gpu case, it exercises gpu branch. In XLA gpu cases, it exercises the default case. This test is to test the non-XLA case so that we disable XLA. We have explicit test for XLA case. PiperOrigin-RevId: 318292680 Change-Id: I5720bab3d98c861951a09b62d09fe9ef0a5abb10
-
由 Hongkun Yu 提交于
PiperOrigin-RevId: 318291000 Change-Id: I2dbd4d3d70b09a08282a5cc42fa5c69de64268da
-
- 25 6月, 2020 21 次提交
-
-
由 Rahul Joshi 提交于
PiperOrigin-RevId: 318279074 Change-Id: I9845b0278737a4d91b0e1e6699ae008d78e76556
-
由 Adrian Kuegel 提交于
PiperOrigin-RevId: 318257426 Change-Id: I83440763f2224463966c40138b5d72f3de741e51
-
https://github.com/llvm/llvm-project/commit/9fb7e98db5aa由 Tres Popp 提交于
PiperOrigin-RevId: 318257184 Change-Id: I5ffc7ac0661dc40615fcfbfdbf4992b035109a50
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 318231045 Change-Id: Idfe29a20335a70aadb152f5612eef8c6d5160b92
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 318231043 Change-Id: Ic8bf82284920a04fe9d1589753905c69bbf8b8e4
-
由 Stephan Herhut 提交于
PiperOrigin-RevId: 318230546 Change-Id: I3e95853baf3e2a8d478565ad495ee1260ef19acb
-
https://github.com/llvm/llvm-project/commit/4c6548222b3c由 Tres Popp 提交于
PiperOrigin-RevId: 318226982 Change-Id: Ib2c98ac9c9f2b10a8f155b512b9844b64acbf46f
-
由 Henry Tan 提交于
PiperOrigin-RevId: 318220378 Change-Id: Ic7b1060986d2e0089177809b234d703387423970
-
由 Henry Tan 提交于
PiperOrigin-RevId: 318217549 Change-Id: I2462eb2b1c294c1adfa88094f714cb2700b5a7d0
-
由 George Karpenkov 提交于
Aliasing for an HLO module is decided at compile time, but the buffer is either donated or not at runtime. In order to avoid recompilations for all possible aliasing configurations it is often advantageous to opportunistically assign an aliasing which may or may not actually hold at runtime. In this case "copy protection" kicks in: if the aliasing specifies that the output buffer B is aliased to input buffer A, but A is not actually donated at runtime, we instead allocate a new fresh buffer C for the output, and copy the contents of A into C (hence "copy protection"). PiperOrigin-RevId: 318214814 Change-Id: Ib5333aafbb5428308e15c18c950110ec6ddecdc5
-
由 Marcello Maggioni 提交于
PiperOrigin-RevId: 318213693 Change-Id: I6a64e9237328388782e481de6ff0268ae11e9695
-
由 A. Unique TensorFlower 提交于
[XLA] Extend implementation of Comparison in HLO, so that the instruction can be more readily extended to support total ordering of all comparison directions. A new comparison type, kFloatTotalOrder, is added but its lowering is not yet supported. Its implementation will be later added to support more precise comparison of floating point numbers. PiperOrigin-RevId: 318213010 Change-Id: I7886a99d2188ea7502e2f00d159492a95d147002
-
由 Christian Sigg 提交于
PiperOrigin-RevId: 318212634 Change-Id: Iada36cd6cf549b4c615be29b76803637de36d5de
-
由 Marcello Maggioni 提交于
Adding an HLO pass that expands the kLogistic HLO into a desired sequence of different HLOs. Currently two different strategies are added. 1) A lowering through an expansion using TAHN (0.5 + 0.5 * tanh(0.5 * x)) 2) A lowering through an expansion using EXP (1.0 / (1.0 + exp(-x))) PiperOrigin-RevId: 318208462 Change-Id: Ibcfba8e95f76c85cdbffc42566f5cec5e663c72b
-
由 Haoyu Zhang 提交于
PiperOrigin-RevId: 318205034 Change-Id: I9dfceae9d9c9896931c99c4f39e0227b8da17df5
-
由 A. Unique TensorFlower 提交于
class. PiperOrigin-RevId: 318203984 Change-Id: I85e2aa5d4c498b67f3e9130fa45a38173f84b35b
-
由 Marcello Maggioni 提交于
It should have a cost similar to kSqrt. PiperOrigin-RevId: 318203675 Change-Id: Ic55aba543bd1760ca0d89c62401524f4ca4ac97b
-
由 Saurabh Saxena 提交于
Enabled unified API test to run with TFRT. Replace dyn_cast with tensorflow::dyn_cast. TF_ExecuteOperation no longer takes a TF_ExecutionContext arg since the current impls tie the op builder to the creating context and it is not clear if we will ever support that API. PiperOrigin-RevId: 318203001 Change-Id: If5048b1404b87c809606c236419e1869630bcd46
-
由 Brian Zhao 提交于
PiperOrigin-RevId: 318199820 Change-Id: I901124780f8687d0f572cff4546f8792f1120e47
-
由 Michael Gester 提交于
Also added required ops and MLIR unit tests. PiperOrigin-RevId: 318199300 Change-Id: I46bf921b5a14c1c4428bfdf51d3e3415a3af65bc
-
由 Saurabh Saxena 提交于
PiperOrigin-RevId: 318190892 Change-Id: I9770cc8689f1ad20c25b04ca2d5503bab9f0b33c
-