MLBEDSW-6748: Update SUPPORTED_OPERATORS.md and release notes
- Updated SUPPORT_OPERATORS.md with Resize operators
- Updated release notes with the main changes and bug fixes
- Updated version numbers
Signed-off-by: oliper01 <oliver.perssonbogdanovski@arm.com>
Signed-off-by: Tim Hall <tim.hall@arm.com>
Change-Id: If25b5fab708098bc3e7eb243924b55a50f148c3a
diff --git a/README.md b/README.md
index 9ed3602..3bd92c8 100644
--- a/README.md
+++ b/README.md
@@ -29,7 +29,8 @@
## TensorFlow Support
-* Vela 3.4.0 to current supports TensorFlow 2.8
+* Vela 3.5.0 to current supports TensorFlow 2.9
+* Vela 3.4.0 supports TensorFlow 2.8
* Vela 3.3.0 supports TensorFlow 2.7
* Vela 3.1.0 to 3.2.0 supports TensorFlow 2.5
* Vela 2.1.0 to 3.0.0 supports TensorFlow 2.4
diff --git a/RELEASES.md b/RELEASES.md
index 97067a0..f930031 100644
--- a/RELEASES.md
+++ b/RELEASES.md
@@ -5,6 +5,27 @@
fixed. The version numbering adheres to the
[semantic versioning](https://semver.org/) scheme.
+## Release 3.5.0 - 23/08/2022
+
+**Main feature changes:**
+
+* New operator support: RESIZE_NEAREST_NEIGHBOR, SHAPE
+* Cascading support for elementwise operators
+* Ethos-U65 performance improvements
+* Upgrade TensorFlow Lite support to version 2.9
+ * Under certain circumstanced the optimised model may not match TensorFlow 2.9
+ for int16x8 CONV2D, TRANSPOSE_CONV2D and FULLY_CONNECTED, see
+ <https://github.com/tensorflow/tensorflow/issues/53763> for more information
+
+**Interface changes:**
+
+* None
+
+**Reported defect fixes:**
+
+* SRAM usage is higher than with old `--pareto-metric BwCycMemBlkH` option for some networks (MLCE-860)
+* Assert caused by CONCATENATION operator with invalid tensor shapes (MLCE-896)
+
## Release 3.4.0 - 24/05/2022
**Main feature changes:**
diff --git a/SUPPORTED_OPS.md b/SUPPORTED_OPS.md
index d16d5f8..6a92e82 100644
--- a/SUPPORTED_OPS.md
+++ b/SUPPORTED_OPS.md
@@ -1,7 +1,7 @@
# Supported Ops
This file was automatically generated by Vela using the `--supported-ops-report` parameter.
-Vela version: `3.4.1.dev3+g5c30971e`
+Vela version: `3.5.0`
This file complies with
[**Gitiles Markdown syntax**](https://github.com/google/gitiles/blob/master/Documentation/markdown.md)
@@ -42,6 +42,7 @@
| RELU_N1_TO_1 | [Generic](#tflite-generic-constraints) |
| RESHAPE | [Generic](#tflite-generic-constraints), [Specific](#tflite-reshape-constraints) |
| RESIZE_BILINEAR | [Generic](#tflite-generic-constraints), [Specific](#tflite-resize_bilinear-constraints) |
+| RESIZE_NEAREST_NEIGHBOR | [Generic](#tflite-generic-constraints), [Specific](#tflite-resize_nearest_neighbor-constraints) |
| SHAPE | [Generic](#tflite-generic-constraints) |
| SLICE | [Generic](#tflite-generic-constraints) |
| SOFTMAX | [Generic](#tflite-generic-constraints), [Specific](#tflite-softmax-constraints) |
@@ -115,6 +116,8 @@
- Axis attribute must be in the range [0, <ofm_dimensions>)
- All Input dimensionalities must match OFM dimensionality
- All Input dimensions must match OFM dimension in all axes except the one defined by the axis attribute
+- All Input dimensions must match OFM dimension in all axes except the one defined by the axis attribute
+- The size of the OFM axis must match the sum of all IFM axis defined by the axis attribute
### TFLite CONV_2D Constraints
@@ -273,8 +276,23 @@
- The width and height of the IFM and OFM must match one of the following criteria:
IFM W and H must both be 1
IFM must match OFM
- OFM W and H must be equal and 2/4/8x IFM -1, if align_corners is True
- OFM W and H must be equal and 2/4/8x IFM, if align_corners is False
+ OFM W and H must be equal and OFM W-1 and H-1 must be 2x/4x/8x IFM W-1 and H-1, if align_corners is True
+ OFM W and H must be equal and OFM W and H must be 2x/4x/8x IFM W and H, if align_corners is False
+- The size tensor must match the output tensor shape
+- Both align_corners and half_pixel_centers can't be True
+- half_pixel_centers are not supported
+
+### TFLite RESIZE_NEAREST_NEIGHBOR Constraints
+
+This is a list of constraints that the RESIZE_NEAREST_NEIGHBOR operator must satisfy in order to be scheduled on the NPU.
+
+- The width and height of the IFM and OFM must match one of the following criteria:
+ IFM W and H must both be 1
+ IFM must match OFM
+ OFM W and H must be equal and OFM W-1 and H-1 must be 2x/4x/8x IFM W-1 and H-1, if align_corners is True
+ OFM W and H must be equal and OFM W and H must be 2x/4x/8x IFM W and H, if align_corners is False
+- The size tensor must match the output tensor shape
+- Both align_corners and half_pixel_centers can't be True
- half_pixel_centers are not supported
### TFLite SOFTMAX Constraints
diff --git a/setup.py b/setup.py
index bf64535..bc5ed0d 100644
--- a/setup.py
+++ b/setup.py
@@ -40,7 +40,7 @@
this_directory = os.path.abspath(os.path.dirname(__file__))
with open(os.path.join(this_directory, "README.md"), encoding="utf-8") as f:
long_description = f.read()
- tag = "3.4.0"
+ tag = "3.5.0"
url = f"https://review.mlplatform.org/plugins/gitiles/ml/ethos-u/ethos-u-vela/+/refs/tags/{tag}/"
# Find all markdown links that match the format: [text](link)
for match, link in re.findall(r"(\[.+?\]\((.+?)\))", long_description):