# CUDA WHEEL
test:linux_cuda_wheel_test_filters --test_tag_filters=gpu,requires-gpu,-no_gpu,-no_oss,-tf_tosa,-oss_excluded,-oss_serial,-benchmark-test,-no_cuda11,-no_oss_py38,-no_oss_py39,-no_oss_py310,-no_oss_py313
test:linux_cuda_wheel_test_filters --build_tag_filters=gpu,requires-gpu,-no_gpu,-no_oss,-tf_tosa,-oss_excluded,-oss_serial,-benchmark-test,-no_cuda11,-no_oss_py38,-no_oss_py39,-no_oss_py310,-no_oss_py313

cuda_comment: >
   From the template it looks like you are installing **TensorFlow** (TF) prebuilt binaries:
      * For TF-GPU - See point 1
      * For TF-CPU - See point 2
   -----------------------------------------------------------------------------------------------
   
   **1. Installing **TensorFlow-GPU** (TF) prebuilt binaries**
   
   
   Make sure you are using compatible TF and CUDA versions.

[pip package](https://www.tensorflow.org/install/pip), to
[enable GPU support](https://www.tensorflow.org/install/gpu), use a
[Docker container](https://www.tensorflow.org/install/docker), and
[build from source](https://www.tensorflow.org/install/source).

To install the current release, which includes support for
[CUDA-enabled GPU cards](https://www.tensorflow.org/install/gpu) *(Ubuntu and
Windows)*:

```
$ pip install tensorflow
```

# Checks TFCI_BAZEL_COMMON_ARGS for "gpu" or "cuda", implying that the test is
# relevant. All of the GPU test machines have CUDA installed via other means,
# so I am not sure how to verify that the dependencies themselves are valid for
# the moment.
if [[ "$TFCI_BAZEL_COMMON_ARGS" =~ gpu|cuda ]]; then
  echo "Checking to make sure tensorflow[and-cuda] is installable..."

        `XNNPACK` delegate automatically when the model has a `fp32` operation.
*   GPU
    *   Allow GPU acceleration starting with internal graph nodes
    *   Experimental support for quantized models with the Android GPU delegate
    *   Add GPU delegate whitelist.
    *   Rename GPU whitelist -> compatibility (list).
    *   Improve GPU compatibility list entries from crash reports.
*   NNAPI
    *   Set default value for

        (environ_cp.get('TF_NEED_ROCM', None) == '1')):
      test_and_build_filters += ['-no_windows_gpu', '-no_gpu']
    else:
      test_and_build_filters.append('-gpu')
  elif is_macos():
    test_and_build_filters += ['-gpu', '-nomac', '-no_mac', '-mac_excluded']
  elif is_linux():
    if ((environ_cp.get('TF_NEED_CUDA', None) == '1') or
        (environ_cp.get('TF_NEED_ROCM', None) == '1')):

  // threadpool of GPU event mgr, as that can trigger more callbacks to be
  // scheduled on that same threadpool, causing a deadlock in cases where the
  // caller of event_mgr->ThenExecute() blocks on the completion of the callback
  // (as in the case of ConstOp kernel creation on GPU, which involves copying a
  // CPU tensor to GPU).

        with:
          api-level: ${{ matrix.api-level }}
          arch: ${{ matrix.api-level == '34' && 'x86_64' || 'x86' }}
          force-avd-creation: false
          emulator-options: -no-window -gpu swiftshader_indirect -noaudio -no-boot-anim -camera-back none
          disable-animations: false
          script: echo "Generated AVD snapshot for caching."

      - name: Run Tests

And then, right after the `yield`, we unload the model. This code will be executed **after** the application **finishes handling requests**, right before the *shutdown*. This could, for example, release resources like memory or a GPU.

/// tip

The `shutdown` would happen when you are **stopping** the application.

Maybe you need to start a new version, or you just got tired of running it. 🤷

///

)

config_setting(
    name = "with_xla_support",
    define_values = {"with_xla_support": "true"},
    visibility = ["//visibility:public"],
)

# By default, XLA GPU is compiled into tensorflow when building with
# --config=cuda even when `with_xla_support` is false. The config setting
# here allows us to override the behavior if needed.
config_setting(
    name = "no_xla_deps_in_cuda",