experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
STT-tensorflow/tensorflow/compiler/mlir/lite/tests
History
TensorFlower Gardener 7a6fa7dbc0 Merge pull request #41790 from lgeiger:fix-matmul-fusion 
PiperOrigin-RevId: 341026586
Change-Id: Idb578a4cf48c82abaad2f811612c4df4c2f2752c
2020-11-06 13:01:02 +00:00
..
debuginfo
Add target environment constraints to more TensorFlow packages and targets.
2020-09-24 22:38:24 -07:00
end2end
Update TFLite converter to export saved model signature def to the tflite schema.
2020-10-15 16:21:43 -07:00
flatbuffer2mlir
Update the remaining unchanged spots, which read builtin code
2020-10-20 19:30:01 -07:00
mlir2exec
Add target environment constraints to more TensorFlow packages and targets.
2020-09-24 22:38:24 -07:00
mlir2flatbuffer
Update TFLite converter to export saved model signature def to the tflite schema.
2020-10-15 16:21:43 -07:00
BUILD
Add target environment constraints to more TensorFlow packages and targets.
2020-09-24 22:38:24 -07:00
canonicalize.mlir
[MLIR] Fix incorrect tfl.while canonicalization
2020-07-06 09:26:30 -07:00
const-fold.mlir
Allow running tests with FILECHECK_OPTS=-enable-var-scope
2020-07-27 12:59:40 -07:00
default_quant_params.mlir
Add the passes to support default range quantization
2020-01-08 17:06:45 -08:00
dilated-conv.mlir
Add method SpaceToBatchNDOp::inferReturnTypes. This infers as much information as possible about output type, given the available argument information. Rank can be inferred when input rank is available, and static shape can be inferred when padding and block_shape arguments are constant.
2020-10-04 13:32:43 -07:00
fuse-tftext.mlir
Update tf.Reshape verifier and builder shape inference to support more operand types and more refined output shapes.
2020-10-15 07:04:09 -07:00
inlining.mlir
Remove --dump-input=fail flag passed into FileCheck in FileCheck tests.
2020-06-10 12:53:21 -07:00
legalize-tf-no-runtime-verification.mlir
Move TF Broadcast op legalization process to the prepare_tf stage
2020-08-11 23:09:21 -07:00
legalize-tf-while.mlir
Drop --dump-input-on-failure from FileCheck commands
2020-06-10 12:00:09 -07:00
legalize-tf.mlir
[tf.lite] Add cumsum mlir conversion & e2e tests.
2020-10-19 20:16:59 -07:00
load-quantization-recipe.mlir
Drop --dump-input-on-failure from FileCheck commands
2020-06-10 12:00:09 -07:00
lower-static-tensor-list.mlir
[MLIR] Extend LowerStaticTensorListPass to handle WhileRegion
2020-07-15 09:13:08 -07:00
ops.mlir
Remove same scale constraint from TFL ReLU ops.
2020-09-11 15:34:46 -07:00
optimize_functional_ops.mlir
[MLIR] Determine function visibility during import
2020-06-12 11:53:36 -07:00
optimize.mlir
Merge pull request #41790 from lgeiger:fix-matmul-fusion
2020-11-06 13:01:02 +00:00
post-quantize.mlir
Fixed a bug where RemoveQuantizationAdapterOps would remove a dequantize which fed a return even if that op had other uses.
2020-06-30 12:29:14 -07:00
prepare-composite-functions-tf.mlir
Convert appropriate composite functions to TFLite custom NMS op.
2020-08-25 09:50:46 -07:00
prepare-quantize-signed.mlir
Expand the min/max range to prevent extremely small scales
2020-07-02 23:37:24 -07:00
prepare-quantize.mlir
Replace instances of "whitelist" with "allowlist" where possible. See Google Developer guidelines at https://developers.google.com/style/word-list#blacklist for more information.
2020-07-08 10:13:43 -07:00
prepare-tf.mlir
Canonicalize FusedBatchNorm op to FusedBatchNormV3
2020-10-15 15:34:10 -07:00
quantize.mlir
Remove the redundant tfl.quantizet/tfl.dequantize pair in the post-quantize
2020-05-28 19:11:52 -07:00
raise-custom-ops.mlir
Add a pass to raise unknown ops to new internal op type custom_tf op.
2020-07-21 12:59:54 -07:00
split-merged-operands.mlir
Add op sanity checks for RNN ops in TFLite
2020-05-27 01:15:21 -07:00
tfl_while_op_licm.mlir
Drop --dump-input-on-failure from FileCheck commands
2020-06-10 12:00:09 -07:00
tfl_while_outline.mlir
Drop --dump-input-on-failure from FileCheck commands
2020-06-10 12:00:09 -07:00
trim-functions-tf.mlir
Replace instances of "whitelist" with "allowlist" where possible. See Google Developer guidelines at https://developers.google.com/style/word-list#blacklist for more information.
2020-07-08 10:13:43 -07:00