exit after tests

tests/ci/accelerate.sh

@@ -1,10 +1,77 @@
 # Description: Accelerate CI script
 
-git clone https://github.com/huggingface/accelerate.git --branch hpu-support --depth 1 && cd accelerate
-pip install -e .[testing] git+https://github.com/HabanaAI/[email protected]
+git clone https://github.com/huggingface/accelerate.git --branch hpu-support --depth 1 && cd accelerate && pip install -e .[testing] git+https://github.com/HabanaAI/[email protected]
 
 RUN_SLOW=1 PT_ENABLE_INT64_SUPPORT=1 make test_cli
+
+if [ $? -ne 0 ]; then
+    exit 1
+fi
+
 RUN_SLOW=1 PT_ENABLE_INT64_SUPPORT=1 PT_HPU_LAZY_MODE=0 make test_core
+
+if [ $? -ne 0 ]; then
+    exit 1
+fi
+
 RUN_SLOW=1 PT_ENABLE_INT64_SUPPORT=1 PT_HPU_LAZY_MODE=0 make test_fsdp
+
+if [ $? -ne 0 ]; then
+    exit 1
+fi
+
 RUN_SLOW=1 PT_ENABLE_INT64_SUPPORT=1 PT_HPU_LAZY_MODE=0 make test_deepspeed
+
+if [ $? -ne 0 ]; then
+    exit 1
+fi
+
 RUN_SLOW=1 PT_ENABLE_INT64_SUPPORT=1 PT_HPU_LAZY_MODE=0 make test_big_modeling
+
+if [ $? -ne 0 ]; then
+    exit 1
+fi
+
+#   - name: Run Accelerate Core tests
+#     working-directory: ./accelerate
+#     run: |
+#       make test_core
+#     env:
+#       PT_ENABLE_INT64_SUPPORT: 1 # for tokenizers
+#       PT_HPU_LAZY_MODE: 0 # for fsdp
+#       RUN_SLOW: 1
+
+#   - name: Run Accelerate CLI tests
+#     working-directory: ./accelerate
+#     run: |
+#       make test_cli
+#     env:
+#       PT_ENABLE_INT64_SUPPORT: 1 # for tokenizers
+#       RUN_SLOW: 1
+
+#   - name: Run Accelerate FSDP tests
+#     working-directory: ./accelerate
+#     run: |
+#       make test_fsdp
+#     env:
+#       PT_ENABLE_INT64_SUPPORT: 1 # for tokenizers
+#       PT_HPU_LAZY_MODE: 0 # for fsdp
+#       RUN_SLOW: 1
+
+#   - name: Run Accelerate DeepSpeed tests
+#     working-directory: ./accelerate
+#     run: |
+#       make test_deepspeed
+#     env:
+#       PT_ENABLE_INT64_SUPPORT: 1 # for tokenizers
+#       PT_HPU_LAZY_MODE: 0 # for cpu offload
+#       RUN_SLOW: 1
+
+#   - name: Run Accelerate Big Modeling tests
+#     working-directory: ./accelerate
+#     run: |
+#       make test_big_modeling
+#     env:
+#       PT_ENABLE_INT64_SUPPORT: 1 # for tokenizers
+#       PT_HPU_LAZY_MODE: 0 # for cpu offload
+#       RUN_SLOW: 1