add tests

.github/workflows/test_llama_end_to_end.yml

@@ -0,0 +1,41 @@
+name: Test Llama End to End
+
+on:
+  workflow_dispatch:
+  pull_request:
+  push:
+    branches:
+      - refactor-stateless_llama.py
+
+jobs:
+  test:
+    strategy:
+      matrix:
+        version: [3.11]
+        os: [xida-cpu-0]
+
+    runs-on: ${{matrix.os}}
+    steps:
+      - name: "Setting up Python"
+        uses: actions/setup-python@75f3110429a8c05be0e1bf360334e4cced2b63fa # v2.3.3
+        with:
+          python-version: ${{matrix.version}}
+
+      - name: "Checkout Code"
+        uses: actions/checkout@v2
+
+      - name: Sync source deps
+        run: |
+          python -m pip install --upgrade pip
+          # Note: We install in three steps in order to satisfy requirements
+          # from non default locations first. Installing the PyTorch CPU
+          # wheels saves multiple minutes and a lot of bandwidth on runner setup.
+          pip install --index-url https://download.pytorch.org/whl/cpu \
+            -r pytorch-cpu-requirements.txt \
+            -r torchvision-requirements.txt          
+          pip install --upgrade -r requirements.txt
+          pip install -e  .[testing]
+
+      - name: Run tests
+        run: |
+          pytest tests/custom_models/stateless_llama

tests/custom_models/stateless_llama/vmfb_comparison_test.py

@@ -0,0 +1,47 @@
+import pytest
+import subprocess
+import os
+import glob
+
+def delete_files(pattern):
+    # Delete files matching the given pattern
+    for file in glob.glob(pattern):
+        os.remove(file)
+
+@pytest.fixture(scope="session", autouse=True)
+def setup_and_teardown():
+    # Setup: Delete existing files
+    delete_files('*.safetensors')
+    delete_files('*.ggml')
+    delete_files('*.vmfb')
+    delete_files('*.mlir')
+    # Yield to the test execution
+    yield
+    # Teardown: Delete files after tests
+    delete_files('*.safetensors')
+    delete_files('*.ggml')
+    delete_files('*.vmfb')
+    delete_files('*.mlir')
+
+@pytest.fixture
+def setup_environment():
+    # Change to the SHARK-Turbine directory
+    os.chdir(os.path.expanduser('~/SHARK-Turbine'))
+    # Ensure that any failure in the commands causes the test to stop
+    subprocess.run('set -e', shell=True, check=True)
+
+def run_command(command):
+    # Run the command and check for successful execution
+    subprocess.run(command, shell=True, check=True)
+
+def test_generate_vmfb(setup_environment):
+    command = 'python python/turbine_models/custom_models/stateless_llama_export_old.py --compile_to=vmfb --hf_model_name="llSourcell/medllama2_7b" --precision=f16 --quantization=int4  --external_weights=safetensors'
+    run_command(command)
+
+def test_generate_quantized_safetensors(setup_environment):
+    command = 'python python/turbine_models/gen_external_params/gen_external_params.py --hf_model_name="llSourcell/medllama2_7b" --precision=f16 --quantization=int4'
+    run_command(command)
+
+def test_run_vmfb_vs_torch_model(setup_environment):
+    command = 'python python/turbine_models/custom_models/stateless_llama.py --run_vmfb --hf_model_name="llSourcell/medllama2_7b" --vmfb_path=medllama2_7b.vmfb --external_weight_file=medllama2_7b_f16_int4.safetensors'
+    run_command(command)