test

Dockerfile

@@ -1,6 +1,3 @@
-# Copyright 2022 MosaicML LLM Foundry authors
-# SPDX-License-Identifier: Apache-2.0
-
 ARG BASE_IMAGE
 FROM $BASE_IMAGE
 
@@ -11,20 +8,39 @@ ARG KEEP_FOUNDRY="false"
 
 ENV TORCH_CUDA_ARCH_LIST="8.0 8.6 8.7 8.9 9.0"
 
+RUN echo "Starting build with BRANCH_NAME=${BRANCH_NAME}, DEP_GROUPS=${DEP_GROUPS}, TE_COMMIT=${TE_COMMIT}, KEEP_FOUNDRY=${KEEP_FOUNDRY}"
+
 # Check for changes in setup.py.
-# If there are changes, the docker cache is invalidated and a fresh pip installation is triggered.
 ADD https://raw.githubusercontent.com/mosaicml/llm-foundry/$BRANCH_NAME/setup.py setup.py
 RUN rm setup.py
 
 # Install TransformerEngine
 RUN NVTE_FRAMEWORK=pytorch CMAKE_BUILD_PARALLEL_LEVEL=4 MAX_JOBS=4 pip install git+https://github.com/NVIDIA/TransformerEngine.git@$TE_COMMIT
 
-# Install and uninstall foundry to cache foundry requirements
+RUN echo "Cloning llm-foundry repository"
 RUN git clone -b $BRANCH_NAME https://github.com/mosaicml/llm-foundry.git
+
+RUN echo "Installing llm-foundry"
 RUN pip install --no-cache-dir "./llm-foundry${DEP_GROUPS}"
 
+RUN echo "KEEP_FOUNDRY value: ${KEEP_FOUNDRY}"
+
 # Conditionally uninstall llm-foundry and remove its directory
 RUN if [ "$KEEP_FOUNDRY" != "true" ]; then \
+      echo "Uninstalling llm-foundry"; \
       pip uninstall -y llm-foundry && \
       rm -rf llm-foundry; \
+    else \
+      echo "Keeping llm-foundry installed"; \
     fi
+
+# Add a final check to see if llm-foundry is installed
+RUN pip list | grep llm-foundry || echo "llm-foundry not found in pip list"
+
+# Print the contents of the directory where llm-foundry was cloned
+RUN ls -la /llm-foundry || echo "llm-foundry directory not found"
+
+# Try to import llm-foundry in Python
+RUN python -c "import llm_foundry; print('llm-foundry successfully imported')" || echo "Failed to import llm-foundry"
+
+RUN echo "Build process completed"