aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
...
* | | | | Implement constant buffer allocation for XLA:GPUGravatar Sanjoy Das2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This CL teaches XLA:GPU to use "normal" buffer assignment for constant instructions. Constant instructions are mapped to a BufferAllocation, like all other instructions, except the storage for this buffer is allocated statically as a global in the generated PTX. This CL does not change how we access the constants -- in IrEmitterUnnested::BuildKernelThunk (used for top level computations) and in HloToIrBindings::EmitBasePointersForHlos (used for nested computations) we bind the kConstant instructions to the llvm::GlobalVariable backing them. So users of constant instructions still access the globals corresponding to the constants directly. However, we no longer emit the constant literals inline. Instead we emit a constant with a zero initializer and then memcpy in the contents of the literal when we load the CUBIN/PTX. This works around compile time issues in LLVM and ptxas caused by large constants. We also populate `BufferAllocations` with the device pointers for the constant globals. This is at least needed for TupleThunk today because TupleThunk wants the addresses for the sub-buffers on the host. I'm not sure if there are other places in XLA:GPU that rely on there being an entry in BufferAllocations for every BufferAllocation. PiperOrigin-RevId: 206243319
* | | | | Fix: When sample_weight_mode is a list/dict set default sample_weight values ↵Gravatar Pavithra Vijay2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | so that we do not require sample_weight to be set during training/eval PiperOrigin-RevId: 206242625
* | | | | Remove tensorflow/opensource_only/Core as it is not used by tensorflow.Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206240947
* | | | | [TF:XLA] Align the two implementations of CanShareOperandBufferWithUser.Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Eventually (when TuplePointsToAnalysis is removed), there will be only one implementation left. Also, use early return instead of else-if to make the code less indented. PiperOrigin-RevId: 206240067
* | | | | Avoid using cuda_fp16.h for fp16 definition.Gravatar Gunhan Gulsoy2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206238991
* | | | | [TF:XLA] Remove unused parameter.Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206237934
* | | | | [tf.data] Switch `batch_and_drop_remainder` to use the fused op.Gravatar Derek Murray2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change also switches `padded_batch_and_drop_remainder` to use the corresponding fused op. PiperOrigin-RevId: 206236616
* | | | | Don't make remote copy call when both send/recv devices are the same.Gravatar Akshay Modi2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206236233
* | | | | Increase the input dimension size from 4 to 6 to address the RetinaNet modelGravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206235660
* | | | | Keras Tensorboard Callback - enable metrics logging in Eager modeGravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206235264
* | | | | Test tf.constant directly (still doesn't work with scalars)Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206224062
* | | | | This change adds an attribute to the _HostCast op to have the same interface ↵Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | as the Cast op. PiperOrigin-RevId: 206218592
* | | | | [XLA] Fix BF16 propagation type adjustmentGravatar Yuanzhong Xu2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | AllUsersConsumeBF16() incorrectly used ValueTypeAfterChange() for the current value being checked, but it should be the original type. Also fusion computation should be adjusted as soon as the fusion root is adjusted. There was also redundant work for while computations. Now removed. PiperOrigin-RevId: 206216822
* | | | | [XLA:GPU] Don't fuse random things into fusions that become a cublas callGravatar Benjamin Kramer2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Only transpose and broadcast are valid. I think this used to work because we didn't emit cublas calls for fused dots until recently. PiperOrigin-RevId: 206213730
* | | | | Full set of unittests for the Eager delegate.Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206211243
* | | | | ResourceVariables shouldn't need twice the memory when initializing.Gravatar Alexandre Passos2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is safe because all ops which write to resource variables check whether there are other outstanding references to the buffer and copy if that's the case. So we can safely reuse the buffer of initializer tensors even in weird cases such as initializing from a constant (which should never be mutated) or using the same tensor to initialize multiple variables. PiperOrigin-RevId: 206211065
* | | | | Delete unused code in CXX11/src/NeuralNetworksGravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206209252
* | | | | Restore tower local variables correctly in init_from_checkpoint.Gravatar Priya Gupta2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206208637
* | | | | PUBLIC: Allow user passing training/evaluation/prediction_hooks from ↵Gravatar Youlong Cheng2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | tf.estimator.EstimatorSpec. PiperOrigin-RevId: 206208119
* | | | | [TF:XLA] Bump open source llvm revision to r338012Gravatar Sanjoy Das2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206204159
* | | | | More changes for TFLite portabilityGravatar Jared Duke2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206202981
* | | | | Add example Unity plugin for the experimental TFLite C API.Gravatar Jared Duke2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Instructions for using the Plugin are contained in the readme @ TensorFlowLitePlugin/README.md PiperOrigin-RevId: 206201614
* | | | | Remove the gen_locally tag which is no more needed.Gravatar Guangda Lai2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206197083
* | | | | shard some estimator testsGravatar Zhenyu Tan2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206196742
* | | | | Boosted trees: Revealing pruning mode as one of the parameters for a gbdt ↵Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | estimator PiperOrigin-RevId: 206193733
* | | | | Ergonomic fixes to tensorflow::dump_graph::*Gravatar Sanjoy Das2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Teach the routines to create the dump directories if they don't exist. I routinely forget to create the dump directory before launching the model. - Print out the file locations on LOG(INFO) and the error messages on LOG(WARNING). PiperOrigin-RevId: 206192822
* | | | | Make model totally defunable.Gravatar Xuechen Li2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206192038
* | | | | remove uncessary variable naming and commentGravatar Zhenyu Tan2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206191743
* | | | | [SE] Try again to query the GPU driver for error descriptionsGravatar Benjamin Kramer2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This code hs been here since 2014, now the oldest supported version of CUDA is 8 so cuGetErrorName should always be available. Also the list of errors is (of course) out of sync with upstream CUDA. Also surface the description of the error to the user, if available. PiperOrigin-RevId: 206191424
* | | | | Replace //tools/defaults:crosstool to ↵Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | @bazel_tools//tools/cpp:current_cc_toolchain //tools/defaults will be removed soon PiperOrigin-RevId: 206187625
* | | | | Add one_hot op support to TFLiteGravatar Jared Duke2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206185190
* | | | | Adding core estimator for a fusion model.Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206183643
* | | | | Don't merge co-located buffer sets containing constantsGravatar Sanjoy Das2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is analogous to cr/188117187, but for constants. PiperOrigin-RevId: 206179508
* | | | | Allow setting server_def on tf.enable_eager_executionGravatar Akshay Modi2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206179345
* | | | | Make function_utils.get_func_code more tolerant of strange objects like ↵Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | functool.partial. PiperOrigin-RevId: 206175973
| | | | * Fix formattingGravatar gracehoney2018-07-26
| | | | |
* | | | | fixing bug for when num of trees is not setGravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206171508
* | | | | Merge pull request #21100 from aaroey:fix_engine_op_fallbackGravatar TensorFlower Gardener2018-07-26
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206169769
* | | | | | Automated rollback of commit b8a9d163d9cbb4b581c044d9c4b1b256c801a9c4Gravatar Russell Power2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206166233
* | | | | | [XLA:GPU] Enable TF RNG ops for the GPU backend.Gravatar Bixia Zheng2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove flag tf_enable_prng_ops_gpu. Pass Tensorflow random number generation ops to the GPU backend. PiperOrigin-RevId: 206165176
* | | | | | remove as much as possible from the try-blocks of contextlib context managers.Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206163905
* | | | | | Update error messages in tflite_diff.Gravatar Nupur Garg2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206162326
* | | | | | A TF Lite kernel to execute TensorFlow delegated ops via Eager.Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206160001
* | | | | | Support sorting of key/value pairs on GPU.Gravatar Adrian Kuegel2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This requires a slight modification in the emitted compare loop: now, we use another if to check if we need to swap instead of two selects. Speed is mostly the same, possibly even a little bit faster. PiperOrigin-RevId: 206148647
* | | | | | Internal change.Gravatar Lasse Espeholt2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206139578
* | | | | | Small changes to placer comments and remove one conditional from Find().Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206125313
* | | | | | Convert device function stack into TraceableStack for use in error message ↵Gravatar A. Unique TensorFlower2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | interpolation. PiperOrigin-RevId: 206120307
* | | | | | Support PREDICT mode for BROADCAST input pipeline.Gravatar Youlong Cheng2018-07-26
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206112531
* | | | | | Ensure failed sub-streams are not re-used.Gravatar Todd Wang2018-07-25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Streams have a monotonic state machine; if a stream encounters any error, it will remain in an error state forever. Without this change, a previously failed sub-stream will be put back on sub_streams_, only to cause the next usage of the sub-stream to trivially fail. PiperOrigin-RevId: 206112024
* | | | | | [XLA] This is a step to incrementally move client/xla_client/* to client/.Gravatar A. Unique TensorFlower2018-07-25
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 206111380
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-09 23:32:37 +0000