mlir-hlo

7b0a6bfeee Integrate LLVM at llvm/llvm-project@fcf6800506 A. Unique TensorFlower 2021-03-31 13:14:22 -0700
af3bc47a8b Integrate LLVM at llvm/llvm-project@8396aeb07c A. Unique TensorFlower 2021-03-31 08:00:52 -0700
bbe0aa204c [MLIR][MHLO] Merge assuming ops with compatible witnesses A. Unique TensorFlower 2021-03-31 06:10:15 -0700
c8157ba4df Move code from helper struct to the only user. Adrian Kuegel 2021-03-31 05:20:37 -0700
4033a56750 Add special cases for SelectOp rank specialization. Adrian Kuegel 2021-03-31 04:28:00 -0700
9206805c58 [MLIR][MHLO] Do not yield results of ops that were moved out of assuming regions A. Unique TensorFlower 2021-03-31 03:49:29 -0700
8ade5d78c8 [MLIR][MHLO] Move `cstr_broadcastable` and `shape_of` out of `assuming` regions A. Unique TensorFlower 2021-03-31 02:38:29 -0700
af2aaa6144 Integrate LLVM at llvm/llvm-project@afed50a14b A. Unique TensorFlower 2021-03-31 01:36:00 -0700
eade942635 [MLIR][MHLO] Add pattern to move ops into the assuming region A. Unique TensorFlower 2021-03-31 01:22:20 -0700
5ec66775d4 Temporarily relax restriction on mhlo.while terminator Geoffrey Martin-Noble 2021-03-30 19:40:48 -0700
5d65758e8c Canonicalize MHLO Case and If Ops with constant conditions Geoffrey Martin-Noble 2021-03-30 17:55:38 -0700
2fb2a92c6e Verify mhlo.if region return types match op Geoffrey Martin-Noble 2021-03-30 17:55:12 -0700
3be9874d82 Remove comments for license declarations. These can get out of date/stale. Russell Power 2021-03-30 17:54:53 -0700
00552afdd3 Integrate LLVM at llvm/llvm-project@c06a8f9caa A. Unique TensorFlower 2021-03-30 17:50:03 -0700
7a9394dca5 Restrict MHLO control flow ops to single-block regions Geoffrey Martin-Noble 2021-03-30 17:48:56 -0700
e78c59d927 Integrate LLVM at llvm/llvm-project@73adc05ced A. Unique TensorFlower 2021-03-30 14:48:05 -0700
3d8095a5ec Integrate LLVM at llvm/llvm-project@77d81c2270 A. Unique TensorFlower 2021-03-30 12:25:51 -0700
f90042321a Integrate LLVM at llvm/llvm-project@c51e91e046 A. Unique TensorFlower 2021-03-30 07:04:30 -0700
c1a6ae8994 Generalize the HloBinaryElementwiseAdaptor Adrian Kuegel 2021-03-30 03:53:04 -0700
6388e8d9ee mlir-hlo-opt: set preloadDialectsInContext to false. Adrian Kuegel 2021-03-30 01:06:12 -0700
9ebadc4c4d Integrate LLVM at llvm/llvm-project@482283042f A. Unique TensorFlower 2021-03-29 18:28:52 -0700
12a82dfe44 Integrate LLVM at llvm/llvm-project@20d5c42e0e A. Unique TensorFlower 2021-03-29 14:35:17 -0700
85a306d356 [MLIR][MHLO] Add pattern to inline broadcasted shapes A. Unique TensorFlower 2021-03-29 06:31:43 -0700
fb819c1de8 [MLIR][MHLO] Apply patterns in MoveUpDynamicBroadcastsForFusionPass greedily A. Unique TensorFlower 2021-03-29 06:00:09 -0700
238c1d8a92 Integrate LLVM at llvm/llvm-project@594e0ba969 A. Unique TensorFlower 2021-03-26 16:43:55 -0700
9b7479f907 Integrate LLVM at llvm/llvm-project@4157a079af A. Unique TensorFlower 2021-03-25 16:38:32 -0700
a2b6060c0c Add folder for HLO NotOp Geoffrey Martin-Noble 2021-03-25 02:07:29 -0700
cbef26c6a8 Integrate LLVM at llvm/llvm-project@1f6a57c1a0 A. Unique TensorFlower 2021-03-24 08:09:47 -0700
52d1744285 Integrate LLVM at llvm/llvm-project@3e0ad11543 A. Unique TensorFlower 2021-03-24 06:43:34 -0700
a34aa699f8 Fix tanh lowering for NaN input. Adrian Kuegel 2021-03-24 06:33:43 -0700
7dd0fe4592 Integrate LLVM at llvm/llvm-project@d709dcc090 A. Unique TensorFlower 2021-03-23 17:57:08 -0700
7f2bf48b8b Integrate LLVM at llvm/llvm-project@b24436ac96 Stella Laurenzo 2021-03-23 12:18:57 -0700
8987dfd1d6 [MLIR][HLO] Move broadcasts over n-ary shape-preserving ops A. Unique TensorFlower 2021-03-23 09:38:00 -0700
5bc4bf0834 Integrate LLVM at llvm/llvm-project@0776eca7a4 A. Unique TensorFlower 2021-03-23 07:25:17 -0700
618223778d Integrate LLVM at llvm/llvm-project@5657f93e78 A. Unique TensorFlower 2021-03-23 06:14:57 -0700
1a6483e229 Integrate LLVM at llvm/llvm-project@e990fa2170 A. Unique TensorFlower 2021-03-23 04:38:28 -0700
54f37abc28 [MHLO] Move broadcasts over elementwise ops A. Unique TensorFlower 2021-03-23 02:33:46 -0700
98debb127d Update Tensorflow to prepare for changing the default value of `preloadDialectsInContext` Mehdi Amini 2021-03-20 11:00:55 -0700
431be0e9b2 Integrate LLVM at llvm/llvm-project@cd442157cf A. Unique TensorFlower 2021-03-19 05:54:15 -0700
59fa7c0ef7 [MHLO:linalg] Lower all dynamic broadcasts of static shapes to linalg.generic Benjamin Kramer 2021-03-19 03:51:14 -0700
22b27dbaa2 Integrate LLVM at llvm/llvm-project@628f5c9da2 A. Unique TensorFlower 2021-03-19 03:45:30 -0700
dc3dbb9d17 Integrate LLVM at llvm/llvm-project@80df56f7f9 A. Unique TensorFlower 2021-03-19 01:34:30 -0700
a26e227b46 Integrate LLVM at llvm/llvm-project@16c30c3c23 A. Unique TensorFlower 2021-03-18 20:15:10 -0700
76d61d4d51 Integrate LLVM at llvm/llvm-project@078b338ba6 A. Unique TensorFlower 2021-03-18 08:24:30 -0700
8ac890d531 Integrate LLVM at llvm/llvm-project@388fbefb4f A. Unique TensorFlower 2021-03-18 06:09:41 -0700
0c4a89e52c [MLIR][MHLO] Implement shape reification for `dynamic_broadcast_in_dim` A. Unique TensorFlower 2021-03-18 03:38:18 -0700
aad6d39952 Integrate LLVM at llvm/llvm-project@0604e0bc68 A. Unique TensorFlower 2021-03-18 00:51:33 -0700
f7464dfa62 Integrate LLVM at llvm/llvm-project@2416f24363 A. Unique TensorFlower 2021-03-17 20:21:20 -0700
f5c9955056 Integrate LLVM at llvm/llvm-project@ecfa874531 A. Unique TensorFlower 2021-03-17 13:31:22 -0700
f1408e791e [MLIR][HLO] Add `Elementwise` trait to unary element-wise ops A. Unique TensorFlower 2021-03-17 08:49:22 -0700
cd52adb20e Integrate LLVM at llvm/llvm-project@967b64beb4 A. Unique TensorFlower 2021-03-17 07:02:19 -0700
2e0ee7759b Add support for lowering mhlo.torch_index_select to Linalg on tensors. Hanhan Wang 2021-03-17 06:31:58 -0700
1336c95920 Integrate LLVM at llvm/llvm-project@506df1bbfd A. Unique TensorFlower 2021-03-17 02:50:06 -0700
a58e62590e Restrict canonicalization to avoid changing type Jacques Pienaar 2021-03-16 16:53:10 -0700
caae2525ef Integrate LLVM at llvm/llvm-project@85ab413b53 A. Unique TensorFlower 2021-03-16 16:40:21 -0700
c54527fe88 Integrate LLVM at llvm/llvm-project@678241795c A. Unique TensorFlower 2021-03-16 13:31:59 -0700
2be112a603 [MLIR][MHLO] Approximate `tf.Tanh` as constant +/-1 for small/large values A. Unique TensorFlower 2021-03-16 10:11:47 -0700
3de2024a9b Avoid creating tuple type only for verification Jacques Pienaar 2021-03-15 17:57:17 -0700
01d729d35d Integrate LLVM at llvm/llvm-project@6878be5dc3 A. Unique TensorFlower 2021-03-15 11:13:16 -0700
570d29d643 Integrate LLVM at llvm/llvm-project@e9e788d145 A. Unique TensorFlower 2021-03-14 16:13:03 -0700
d16860d26d [MLIR] Change LMHLO Conditional and While to capture needed buffers, instead of passing them by operands. Tim Shen 2021-03-11 14:41:50 -0800
8066794eea Integrate LLVM at llvm/llvm-project@720a828045 A. Unique TensorFlower 2021-03-11 14:17:16 -0800
4f5e1c51dd Add support for lowering NHWC pooling mhlo.reduce_window to Linalg on tensors. Hanhan Wang 2021-03-11 09:40:31 -0800
630cabefb0 Add support for lowering 2D depthwise mhlo.conv to Linalg on tensors. Hanhan Wang 2021-03-11 08:40:31 -0800
94f9740c67 [MLIR][HLO:Linalg] Lower mhlo.dynamic_iota to indexed_generic Benjamin Kramer 2021-03-11 08:30:08 -0800
09f8046816 [MLIR:HLO:LINALG] Fix codegen for mhlo.reshape when one side is rank 0 Benjamin Kramer 2021-03-11 05:28:51 -0800
d77b556822 [MLIR][MHLO] Allow recursion in the shape_of mover Benjamin Kramer 2021-03-11 02:51:27 -0800
67a770e4e0 [HLO:MLIR] Make binary op type reification emit shape_of instead of tensor ops Benjamin Kramer 2021-03-11 02:00:50 -0800
9902e6ee32 [HLO] Add LMHLO CollectivePermute verification. Rahul Joshi 2021-03-10 15:36:22 -0800
4f16b10ce2 Integrate LLVM at llvm/llvm-project@4c973ae51b A. Unique TensorFlower 2021-03-10 12:34:07 -0800
b212bd66ae Build fix for missing precision_config. Mahesh Ravishankar 2021-03-10 11:09:41 -0800
e199df1dbf [MLIR][MHLO] Declare `shape_of` dynamically legal in move-up-dynamic-broadcasts A. Unique TensorFlower 2021-03-10 09:58:15 -0800
c217a6ef61 [MHLO] Add pass to move up dynamic broadcasts for fusion A. Unique TensorFlower 2021-03-10 06:20:43 -0800
cabd4d9a06 Canonicalize dynamic_broadcast_in_dim to own shape with rank narrowing on the shape to a corresponding tensor.cast. Stephan Herhut 2021-03-10 05:43:10 -0800
507d9fb61d [MLIR][KernelGen] Add `tf.Polygamma` kernel A. Unique TensorFlower 2021-03-10 02:20:45 -0800
218476128e [MLIR][KernelGen] Fix zeta lowering at poles A. Unique TensorFlower 2021-03-10 01:08:02 -0800
7629dfdd81 Integrate LLVM at llvm/llvm-project@df6d0579e1 A. Unique TensorFlower 2021-03-09 11:25:59 -0800
5be8be31b5 Integrate LLVM at llvm/llvm-project@3f3f88fb95 Benjamin Kramer 2021-03-09 02:18:06 -0800
daf6bde6f5 Integrate LLVM at llvm/llvm-project@c9ff39a3f9 A. Unique TensorFlower 2021-03-08 14:17:09 -0800
55eda81407 [MLIR][HLO] Reify shape extents as `index` values A. Unique TensorFlower 2021-03-08 02:41:10 -0800
5a415de33b Integrate LLVM at llvm/llvm-project@9b302513f6 Benjamin Kramer 2021-03-05 04:47:46 -0800
d5f80f0469 Integrate LLVM at llvm/llvm-project@cedc53254a A. Unique TensorFlower 2021-03-05 00:36:43 -0800
29f70cb892 PR #46723: Adjust types of loop counters Marius Brehler 2021-03-04 07:35:15 -0800
57e9941d5d Integrate LLVM at llvm/llvm-project@b3a33553ae Benjamin Kramer 2021-03-04 07:22:16 -0800
39650a5d5a Remove rank 1 specialization from TransformUnrankedHloPass. A. Unique TensorFlower 2021-03-04 05:23:52 -0800
e5a6706260 Integrate LLVM at llvm/llvm-project@c907681b07 Benjamin Kramer 2021-03-04 05:21:32 -0800
62b357b601 Remove rank 1 specialization from TransformUnrankedHloPass. Adrian Kuegel 2021-03-04 04:02:50 -0800
50a516fb9c Adopt td_library Geoffrey Martin-Noble 2021-03-03 16:10:20 -0800
5eac983723 Integrate LLVM at llvm/llvm-project@5d7e0a23c6 Benjamin Kramer 2021-03-03 11:07:31 -0800
ab8bc35efd Integrate LLVM at llvm/llvm-project@8da090381d Benjamin Kramer 2021-03-03 09:09:39 -0800
bf14340316 Integrate LLVM at llvm/llvm-project@1a4990a4f7 Benjamin Kramer 2021-03-03 04:55:23 -0800
24c98d5211 Integrate LLVM at llvm/llvm-project@99a6d003ed A. Unique TensorFlower 2021-03-02 21:47:17 -0800
8687f3e4cf Lower MHLO Dot to type-polymorphic linalg named ops Geoffrey Martin-Noble 2021-03-02 13:59:50 -0800
1facbe9eb5 Integrate LLVM at llvm/llvm-project@7f086d74c3 Benjamin Kramer 2021-03-02 08:32:05 -0800
0683db3b24 Legalize MinimumBroadcastShapes op. Adrian Kuegel 2021-03-02 06:37:48 -0800
329b1fd071 Verify compatible shapes in unpack verification rather than exact Jacques Pienaar 2021-03-01 07:59:26 -0800
70ee9369d5 Use mlir::OpState::operator->() to get to Operation::getAttrs(). Christian Sigg 2021-03-01 04:51:42 -0800
7c071e8ee6 Integrate LLVM at llvm/llvm-project@99c24f7aa8 Benjamin Kramer 2021-03-01 03:43:37 -0800
e19ccf975e Filter static dimensions from dynamic_broadcast_in_dim's init_tensor Benjamin Kramer 2021-03-01 03:03:03 -0800
e6a1f5f0f9 Add MinimumBroadcastShapesOp to chlo dialect. Adrian Kuegel 2021-03-01 02:22:55 -0800

Commit Graph Select branches Hide Pull Requests master Mono Color

Commit Graph

Select branches

Hide Pull Requests

master