About updating to tensorflow-icg based on TF version >= 2.0 #6
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
PiperOrigin-RevId: 264736469
… will use new converters. PiperOrigin-RevId: 264736807
This CL makes use of the standard LLVM LLJIT and removes the need for a custom JIT implementation within MLIR. To achieve this, one needs to clone (i.e. serde) the produced llvm::Module into a new LLVMContext. This is currently necessary because the llvm::LLVMContext is owned by the LLVMDialect, somewhat deep in the call hierarchy. In the future we should remove the reliance of serding the llvm::Module by allowing the injection of an LLVMContext from the top-level. Unfortunately this will require deeper API changes and impact multiple places. It is therefore left for future work. PiperOrigin-RevId: 264737459
This CL adds the necessary plumbing to have a SavedModel importer: creating a subclass from ImporterBase, adding entry point functions, and registering it to MLIR translate framework. Right now the importer just converts all functions in the SavedModel. Adding support for SignatureDef and other SavedModel components is to be implemented in subsequent CLs. Added a minimal test to make sure the plumbing works. PiperOrigin-RevId: 264738045
…ialect.cpp - NFC PiperOrigin-RevId: 264740014
PiperOrigin-RevId: 264742130
Fixes #97 PiperOrigin-RevId: 264743395
PiperOrigin-RevId: 264744148
PiperOrigin-RevId: 264744295
…at are present in MatMul Op. In preparation for tf.einsum switch, these are all needed for existing tests to pass. PiperOrigin-RevId: 264744322
…r TF to TFLite translation. PiperOrigin-RevId: 264746751
Operation interfaces generally require a bit of boilerplate code to connect all of the pieces together. This cl introduces mechanisms in the ODS to allow for generating operation interfaces via the 'OpInterface' class.
Providing a definition of the `OpInterface` class will auto-generate the c++
classes for the interface. An `OpInterface` includes a name, for the c++ class,
along with a list of interface methods. There are two types of methods that can be used with an interface, `InterfaceMethod` and `StaticInterfaceMethod`. They are both comprised of the same core components, with the distinction that `StaticInterfaceMethod` models a static method on the derived operation.
An `InterfaceMethod` is comprised of the following components:
* ReturnType
- A string corresponding to the c++ return type of the method.
* MethodName
- A string corresponding to the desired name of the method.
* Arguments
- A dag of strings that correspond to a c++ type and variable name
respectively.
* MethodBody (Optional)
- An optional explicit implementation of the interface method.
def MyInterface : OpInterface<"MyInterface"> {
let methods = [
// A simple non-static method with no inputs.
InterfaceMethod<"unsigned", "foo">,
// A new non-static method accepting an input argument.
InterfaceMethod<"Value *", "bar", (ins "unsigned":$i)>,
// Query a static property of the derived operation.
StaticInterfaceMethod<"unsigned", "fooStatic">,
// Provide the definition of a static interface method.
// Note: `ConcreteOp` corresponds to the derived operation typename.
StaticInterfaceMethod<"Operation *", "create",
(ins "OpBuilder &":$builder, "Location":$loc), [{
return builder.create<ConcreteOp>(loc);
}]>,
// Provide a definition of the non-static method.
// Note: `op` corresponds to the derived operation variable.
InterfaceMethod<"unsigned", "getNumInputsAndOutputs", (ins), [{
return op.getNumInputs() + op.getNumOutputs();
}]>,
];
PiperOrigin-RevId: 264754898
…ast subgraph PiperOrigin-RevId: 264755031
PiperOrigin-RevId: 264755855
PiperOrigin-RevId: 264758470
… max for quantized int32. PiperOrigin-RevId: 264765625
PiperOrigin-RevId: 264780793
The arguments to ComputePaddingHeightWidth were mixed up: dilation rate is 2d (height, width), padding is only passed in once. PiperOrigin-RevId: 264787666
PiperOrigin-RevId: 264789669
PiperOrigin-RevId: 264789677
PiperOrigin-RevId: 264790790
PiperOrigin-RevId: 264802821
PiperOrigin-RevId: 264805165
MLIR builders are meant to be trivially copyable values. PiperOrigin-RevId: 264807495
This delegates handling of custom calls to the extracted ThunkEmitter and adds basic data structures to the HLODialectEmitter to track thunks. PiperOrigin-RevId: 264808440
PiperOrigin-RevId: 264824470
PiperOrigin-RevId: 264826413
…e it. This avoids a potentially quadratic execution time in building the gradient graph, because we were previously creating the set multiple times for each op in the graph. PiperOrigin-RevId: 264826531
PiperOrigin-RevId: 264827908
…ased once a Toeplitz solver is checked in. PiperOrigin-RevId: 264829158
PiperOrigin-RevId: 265389130
PiperOrigin-RevId: 265395319
PiperOrigin-RevId: 265398643
… tflite fused ops. PiperOrigin-RevId: 265400666
PiperOrigin-RevId: 265411691
PiperOrigin-RevId: 265411696
…rouped_conv PiperOrigin-RevId: 265428511
PiperOrigin-RevId: 265440617
This is a part of a larger migration effort for tensorflow::tstring. See: tensorflow/community#91 PiperOrigin-RevId: 265449545
Added Cord copy-ctor/assignment for tstring. This is a part of a larger migration effort for tensorflow::tstring. See: tensorflow/community#91 PiperOrigin-RevId: 265451696
…non-ARM. PiperOrigin-RevId: 265452203
PiperOrigin-RevId: 265455751
PiperOrigin-RevId: 265455830
PiperOrigin-RevId: 265458112
…ing. This is a part of a larger migration effort for tensorflow::tstring. See: tensorflow/community#91 PiperOrigin-RevId: 265474945
This is a part of a larger migration effort for tensorflow::tstring. See: tensorflow/community#91 PiperOrigin-RevId: 265475150
This fixes a bug when folding ops with inner ops and inner ops are still being visited. PiperOrigin-RevId: 265475780
PiperOrigin-RevId: 265479073
PiperOrigin-RevId: 265483372
The `Graph.create_op()` method has a deprecated argument wrapper and performs redundant type checking. Both add unnecessary overhead for calls from framework-internal sources, so this change switches `tf.constant()` and `OpDefLibrary._apply_op_helper()` to use the internal version. As a consequence, classes that override `Graph.create_op()` must now override `Graph._create_op_internal()`, and this change updates all known classes to do so. PiperOrigin-RevId: 265484157
Each `tf.Tensor` caches a SWIG-wrapped `TF_Output` object to map that tensor to the C API's graph. We currently lazily cache this object in `tf.Tensor` on its first consumption, but it turns out that we already build the object for all tensors as a by-product of calculating tensor types. This change saves the original object in each `tf.Tensor` when it is constructed, which decreases the overhead of graph construction. PiperOrigin-RevId: 265484399
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.
No description provided.