test llama

tenstorrent · Dec 20, 2024 · c215bd6 · c215bd6
1 parent 77ef35c
commit c215bd6
Show file tree

Hide file tree

Showing 2 changed files with 13 additions and 17 deletions.
diff --git a/.github/workflows/build-and-test.yml b/.github/workflows/build-and-test.yml
@@ -43,7 +43,7 @@ jobs:
       matrix:
         build:
           - runs-on: runner
-        test_group_id: [1,2]
+        # test_group_id: [1,2]
 
     runs-on:
       - in-service
@@ -130,11 +130,11 @@ jobs:
         -DCMAKE_CXX_COMPILER_LAUNCHER=ccache
         cmake --build ${{ steps.strings.outputs.build-output-dir }}
 
-    - name: Run Unit Tests
-      shell: bash
-      run: |
-        source env/activate
-        cmake --build ${{ steps.strings.outputs.build-output-dir }} -- run_unit_tests
+    # - name: Run Unit Tests
+    #   shell: bash
+    #   run: |
+    #     source env/activate
+    #     cmake --build ${{ steps.strings.outputs.build-output-dir }} -- run_unit_tests
 
     - name: Run Test
       env:
@@ -147,12 +147,7 @@ jobs:
         apt install -y libgl1 libglx-mesa0
         set -o pipefail # Ensures that the exit code reflects the first command that fails
         pip install pytest-split
-        pytest -m push --splits 2 \
-               --group ${{ matrix.test_group_id }} \
-               --splitting-algorithm least_duration \
-               -m "${{ inputs.test_mark }}" \
-               --junit-xml=${{ steps.strings.outputs.test_report_path }} \
-               2>&1 | tee pytest.log
+        pytest -m forge/test/mlir/llama/test_llama_inference.py
 
     - name: Upload Test Log
       uses: actions/upload-artifact@v4

diff --git a/forge/test/mlir/llama/test_llama_inference.py b/forge/test/mlir/llama/test_llama_inference.py
@@ -10,12 +10,13 @@
 from test.mlir.llama.utils.utils import load_model
 
 
-@pytest.mark.nightly
-@pytest.mark.xfail()
-@pytest.mark.parametrize("model_path", ["openlm-research/open_llama_3b", "meta-llama/Llama-3.2-1B"])
+# @pytest.mark.nightly
+# @pytest.mark.xfail()
+@pytest.mark.push()
+@pytest.mark.parametrize("model_path", ["meta-llama/Llama-3.2-1B"])
 def test_llama_inference(model_path):
-    if model_path == "meta-llama/Llama-3.2-1B":
-        pytest.skip("Skipping test for Llama-3.2-1B model, waiting for new transformers version.")
+    # if model_path == "meta-llama/Llama-3.2-1B":
+    # pytest.skip("Skipping test for Llama-3.2-1B model, waiting for new transformers version.")
 
     # Load Model and Tokenizer
     framework_model, tokenizer = load_model(model_path)