Add preprocess to ci (#6544)

* Add preprocess to ci

[ghstack-poisoned]

* Update on "Add preprocess to ci"


1. Run `test_preprocess.py` in ci.
2. Refactor `test_preprocess.py` by creating the model artifacts (eager, exported, et) once and then re-using them on different inputs.

Test Plan:
```
python -m unittest examples.models.llama3_2_vision.preprocess.test_preprocess
```

[ghstack-poisoned]

* Update on "Add preprocess to ci"


1. Run `test_preprocess.py` in ci.
2. Refactor `test_preprocess.py` by creating the model artifacts (eager, exported, et) once and then re-using them on different inputs.

Test Plan:
```
python -m unittest examples.models.llama3_2_vision.preprocess.test_preprocess
```

[ghstack-poisoned]

* Update on "Add preprocess to ci"


1. Run `test_preprocess.py` in ci.
2. Refactor `test_preprocess.py` by creating the model artifacts (eager, exported, et) once and re-use them on different inputs (previously, create model artifacts for each test).

Test Plan:
```
python -m unittest examples.models.llama3_2_vision.preprocess.test_preprocess
```

[ghstack-poisoned]

* Update on "Add preprocess to ci"


1. Run `test_preprocess.py` in ci.
2. Refactor `test_preprocess.py` by creating the model artifacts (eager, exported, et) once and re-use them on different inputs (previously, create model artifacts for each test).

Test Plan:
```
python -m unittest examples.models.llama3_2_vision.preprocess.test_preprocess
```

[ghstack-poisoned]

* Update on "Add preprocess to ci"


1. Run `test_preprocess.py` in ci.
2. Refactor `test_preprocess.py` by creating the model artifacts (eager, exported, et) once and re-use them on different inputs (previously, create model artifacts for each test).

Test Plan:
```
pytest -c /dev/null -v -n auto examples/models/llama3_2_vision/preprocess/
```

[ghstack-poisoned]

* Update on "Add preprocess to ci"


1. Run `test_preprocess.py` in ci.
2. Refactor `test_preprocess.py` by creating the model artifacts (eager, exported, et) once and re-use them on different inputs (previously, create model artifacts for each test).

Test Plan:
```
pytest -c /dev/null -v -n auto examples/models/llama3_2_vision/preprocess/
```

[ghstack-poisoned]

Author: lucylq

.github/workflows/_unittest.yml

@@ -37,6 +37,9 @@ jobs:
         CMAKE_ARGS="-DEXECUTORCH_BUILD_XNNPACK=ON -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON" \
         .ci/scripts/setup-linux.sh cmake
 
+        # Install llama3_2_vision dependencies.
+        PYTHON_EXECUTABLE=python ./examples/models/llama3_2_vision/install_requirements.sh
+
         # Run pytest with coverage
         pytest -n auto --cov=./ --cov-report=xml
         # Run gtest
@@ -67,6 +70,10 @@ jobs:
         ${CONDA_RUN} --no-capture-output \
         .ci/scripts/setup-macos.sh cmake
 
+        # Install llama3_2_vision dependencies.
+        PYTHON_EXECUTABLE=python ${CONDA_RUN} --no-capture-output \
+        ./examples/models/llama3_2_vision/install_requirements.sh
+
         # Run pytest with coverage
         ${CONDA_RUN} pytest -n auto --cov=./ --cov-report=xml
         # Run gtest

examples/models/llama3_2_vision/preprocess/model.py

@@ -26,6 +26,9 @@ class PreprocessConfig:
     max_num_tiles: int = 4
     tile_size: int = 224
     antialias: bool = False
+    # Used for reference eager model from torchtune.
+    resize_to_max_canvas: bool = False
+    possible_resolutions: Optional[List[Tuple[int, int]]] = None
 
 
 class CLIPImageTransformModel(EagerModelBase):