Upgrade TF version from 2.7 to 2.11 (#749)

* Upgrade bazel version to 5.1.0 for building with TFv2.11.0

* Upgrade to TF v2.11

* Bump up the version of numpy to v1.24.2

* Format

* Fix tf.Variable inside tf.while_loop issues inside optimizers

* Format optimizers

* Upgrade the latest gym for notebook tutorial.

* Fix tutorial error

* Fix tutorial errors

* Upgrade seaborn version to 0.12.0

* Fix typo

* Remove debug print

* Update ci_install bazel version, removed duplicated code lines.

Author: jaeyoo

.bazelversion

@@ -1 +1 @@
-3.7.2
+5.1.0

WORKSPACE

@@ -3,8 +3,8 @@
 
 load("@bazel_tools//tools/build_defs/repo:http.bzl", "http_archive")
 
-EIGEN_COMMIT = "12e8d57108c50d8a63605c6eb0144c838c128337"
-EIGEN_SHA256 = "f689246e342c3955af48d26ce74ac34d21b579a00675c341721a735937919b02"
+EIGEN_COMMIT = "3bb6a48d8c171cf20b5f8e48bfb4e424fbd4f79e"
+EIGEN_SHA256 = "eca9847b3fe6249e0234a342b78f73feec07d29f534e914ba5f920f3e09383a3"
 
 
 http_archive(
@@ -33,10 +33,10 @@ http_archive(
 
 http_archive(
     name = "org_tensorflow",
-    sha256 = "249b48ddee927801c7a4f8e5442cf1a3c860f6f46b85a2ff7a78b501507dd561",
-    strip_prefix = "tensorflow-2.7.0",
+    sha256 = "e52cda3bae45f0ae0fccd4055e9fa29892b414f70e2df94df9a3a10319c75fff",
+    strip_prefix = "tensorflow-2.11.0",
     urls = [
-        "https://github.com/tensorflow/tensorflow/archive/refs/tags/v2.7.0.zip",
+        "https://github.com/tensorflow/tensorflow/archive/refs/tags/v2.11.0.zip",
     ],
 )
 

benchmarks/README.md

@@ -24,7 +24,7 @@ Some notes on benchmark configuration:
 
 For example, to benchmark a dense depth-10 Clifford circuit over 5 qubits call:
 ```
-bazel run -c opt --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=0" --cxxopt="-msse2" \
+bazel run -c opt --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=1" --cxxopt="-msse2" \
   --cxxopt="-msse3" --cxxopt="-msse4" \
   benchmarks/scripts:benchmark_clifford_circuit -- \
   --n_moments 5 --n_qubits 4 \
@@ -39,7 +39,7 @@ benchmarks/scripts/reports/CliffordBenchmarks.benchmark_clifford_circuit_4_5_1
 To benchmark the parameter shift differentiation method on a random depth-10 4-qubit circuit with 10 parameters call, where the circuit will be differentiated
 over 50 trials, each time over a batch of 10 circuits.
 ```
-bazel run -c opt --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=0" --cxxopt="-msse2" \
+bazel run -c opt --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=1" --cxxopt="-msse2" \
   --cxxopt="-msse3" --cxxopt="-msse4" \
   benchmarks/scripts:benchmark_op_gradients -- \
   --n_moments 10 --n_qubits 4 --n_symbols 10 \

configure.sh

@@ -99,13 +99,12 @@ if [[ "$PIP_MANYLINUX2010" == "0" ]]; then
   write_to_bazelrc "build:cuda --crosstool_top=@local_config_cuda//crosstool:toolchain"
 fi
 
-
+write_to_bazelrc "build --experimental_repo_remote_exec"
 write_to_bazelrc "build --spawn_strategy=standalone"
 write_to_bazelrc "build --strategy=Genrule=standalone"
-write_to_bazelrc "build --experimental_repo_remote_exec"
 write_to_bazelrc "build -c opt"
-write_to_bazelrc "build --cxxopt=\"-D_GLIBCXX_USE_CXX11_ABI=0\""
-write_to_bazelrc "build --cxxopt=\"-std=c++14\""
+write_to_bazelrc "build --cxxopt=\"-D_GLIBCXX_USE_CXX11_ABI=1\""
+write_to_bazelrc "build --cxxopt=\"-std=c++17\""
 
 
 if is_windows; then

docs/install.md

@@ -84,21 +84,21 @@ As noted in the TensorFlow
 guide, the <a href="https://bazel.build/" class="external">Bazel</a>
 build system will be required.
 
-Our latest source builds use TensorFlow 2.7.0. To ensure compatibility we use `bazel` version 3.7.2. To remove any existing version of Bazel:
+Our latest source builds use TensorFlow 2.11.0. To ensure compatibility we use `bazel` version 5.1.0. To remove any existing version of Bazel:
 
 <!-- common_typos_disable -->
 <pre class="devsite-click-to-copy">
   <code class="devsite-terminal">sudo apt-get remove bazel</code>
 </pre>
 <!-- common_typos_enable -->
 
-Download and install `bazel` version 3.7.2:
+Download and install `bazel` version 5.1.0:
 
 <!-- common_typos_disable -->
 <pre class="devsite-click-to-copy">
-  <code class="devsite-terminal">wget https://github.com/bazelbuild/bazel/releases/download/3.7.2/bazel_3.7.2-linux-x86_64.deb
+  <code class="devsite-terminal">wget https://github.com/bazelbuild/bazel/releases/download/5.1.0/bazel_5.1.0-linux-x86_64.deb
 </code>
-  <code class="devsite-terminal">sudo dpkg -i bazel_3.7.2-linux-x86_64.deb</code>
+  <code class="devsite-terminal">sudo dpkg -i bazel_5.1.0-linux-x86_64.deb</code>
 </pre>
 <!-- common_typos_enable -->
 
@@ -122,7 +122,7 @@ Finally, confirm installation of the correct `bazel` version:
 ### 4. Build TensorFlow from source
 
 Here we adapt instructions from the TensorFlow [build from source](https://www.tensorflow.org/install/source)
-guide, see the link for further details. TensorFlow Quantum is compatible with TensorFlow version&nbsp;2.7.0.
+guide, see the link for further details. TensorFlow Quantum is compatible with TensorFlow version&nbsp;2.11.0.
 
 Download the
 <a href="https://github.com/tensorflow/tensorflow" class="external">TensorFlow source code</a>:
@@ -131,7 +131,7 @@ Download the
 <pre class="devsite-click-to-copy">
   <code class="devsite-terminal">git clone https://github.com/tensorflow/tensorflow.git</code>
   <code class="devsite-terminal">cd tensorflow</code>
-  <code class="devsite-terminal">git checkout v2.7.0</code>
+  <code class="devsite-terminal">git checkout v2.11.0</code>
 </pre>
 
 Be sure the virtual environment you created in step 2 is activated. Then, install the TensorFlow dependencies:
@@ -141,7 +141,8 @@ Be sure the virtual environment you created in step 2 is activated. Then, instal
   <code class="devsite-terminal">pip install -U pip six numpy wheel setuptools mock 'future>=0.17.1'</code>
   <code class="devsite-terminal">pip install -U keras_applications --no-deps</code>
   <code class="devsite-terminal">pip install -U keras_preprocessing --no-deps</code>
-  <code class="devsite-terminal">pip install numpy==1.19.5</code>
+  <code class="devsite-terminal">pip install numpy==1.24.2</code>
+  <code class="devsite-terminal">pip install packaging requests</code>
 </pre>
 <!-- common_typos_enable -->
 
@@ -153,11 +154,11 @@ Configure the TensorFlow build. When asked for the Python interpreter and librar
 </pre>
 <!-- common_typos_enable -->
 
-Build the TensorFlow package:
+Build the TensorFlow package (Since TF v2.8, `_GLIBCXX_USE_CXX11_ABI` is set to 1, and the c++ codes are all compiled with `-std=c++17`):
 
 <!-- common_typos_disable -->
 <pre class="devsite-click-to-copy">
-  <code class="devsite-terminal">bazel build -c opt --cxxopt="-O3" --cxxopt="-march=native" --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=0" //tensorflow/tools/pip_package:build_pip_package</code>
+  <code class="devsite-terminal">bazel build -c opt --cxxopt="-O3" --cxxopt="-march=native" --cxxopt="-std=c++17" --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=1" //tensorflow/tools/pip_package:build_pip_package</code>
 </pre>
 <!-- common_typos_enable -->
 
@@ -193,7 +194,7 @@ Build the TensorFlow Quantum pip package and install:
 <!-- common_typos_disable -->
 <pre class="devsite-click-to-copy">
   <code class="devsite-terminal">./configure.sh</code>
-  <code class="devsite-terminal">bazel build -c opt --cxxopt="-O3" --cxxopt="-march=native" --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=0" release:build_pip_package</code>
+  <code class="devsite-terminal">bazel build -c opt --cxxopt="-O3" --cxxopt="-march=native" --cxxopt="-std=c++17" --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=1" release:build_pip_package</code>
   <code class="devsite-terminal">bazel-bin/release/build_pip_package /tmp/tfquantum/</code>
   <code class="devsite-terminal">python3 -m pip install /tmp/tfquantum/<var>name_of_generated_wheel</var>.whl</code>
 </pre>

docs/tutorials/barren_plateaus.ipynb

@@ -12,7 +12,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 0,
+      "execution_count": null,
       "metadata": {
         "cellView": "form",
         "colab": {},
@@ -89,7 +89,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 0,
+      "execution_count": null,
       "metadata": {
         "colab": {},
         "colab_type": "code",
@@ -112,7 +112,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 0,
+      "execution_count": null,
       "metadata": {
         "colab": {},
         "colab_type": "code",
@@ -125,7 +125,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 0,
+      "execution_count": null,
       "metadata": {
         "colab": {},
         "colab_type": "code",
@@ -150,7 +150,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 0,
+      "execution_count": null,
       "metadata": {
         "colab": {},
         "colab_type": "code",
@@ -202,7 +202,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 0,
+      "execution_count": null,
       "metadata": {
         "colab": {},
         "colab_type": "code",
@@ -278,7 +278,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 0,
+      "execution_count": null,
       "metadata": {
         "colab": {},
         "colab_type": "code",
@@ -327,7 +327,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 0,
+      "execution_count": null,
       "metadata": {
         "colab": {},
         "colab_type": "code",
@@ -399,7 +399,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 0,
+      "execution_count": null,
       "metadata": {
         "colab": {},
         "colab_type": "code",
@@ -455,7 +455,7 @@
     },
     {
       "cell_type": "code",
-      "execution_count": 0,
+      "execution_count": null,
       "metadata": {
         "colab": {},
         "colab_type": "code",
@@ -511,6 +511,15 @@
       "display_name": "Python 3",
       "language": "python",
       "name": "python3"
+    },
+    "language_info": {
+      "name": "python",
+      "version": "3.10.9 (main, Dec  7 2022, 13:47:07) [GCC 12.2.0]"
+    },
+    "vscode": {
+      "interpreter": {
+        "hash": "916dbcbb3f70747c44a77c7bcd40155683ae19c65e1c03b4aa3499c5328201f1"
+      }
     }
   },
   "nbformat": 4,