fixes

Author: metascroy

.github/workflows/pull.yml

@@ -738,68 +738,68 @@ jobs:
       matrix:
         runner: [macos-m1-stable ]
     runs-on: ${{matrix.runner}}
-      steps:
-        - name: Setup Python
-          uses: actions/setup-python@v2
-          with:
-            python-version: 3.10.11
-        - name: Run test
-          run: |
-            export PYTHON_VERSION="3.10"
-            set -x
-            # NS/MC: Remove previous installation of torch and torchao first
-            # as this script does not install anything into conda env but rather as system dep
-            pip3 uninstall -y torch || true
-            set -eou pipefail
-
-            pip3 uninstall -y torchao || true
-            set -eou pipefail
-
-            echo "::group::Print machine info"
-            uname -a
-            sysctl machdep.cpu.brand_string
-            sysctl machdep.cpu.core_count
-            echo "::endgroup::"
-
-            echo "::group::Install requirements"
-            # Install requirements
-            ./install/install_requirements.sh
-            ls -la
-            pwd
-            pip3 list
-            python3 -c 'import torch;print(f"torch: {torch.__version__, torch.version.git_version}")'
-            echo "::endgroup::"
-
-            echo "::group::Download checkpoints"
-            (
-              mkdir -p checkpoints/stories15M
-              pushd checkpoints/stories15M
-              curl -fsSL -O https://huggingface.co/karpathy/tinyllamas/resolve/main/stories15M.pt
-              curl -fsSL -O https://github.com/karpathy/llama2.c/raw/master/tokenizer.model
-              popd
-            )
-            echo "::endgroup::"
-
-            echo "::group::Run inference"
-            export MODEL_PATH=checkpoints/stories15M/stories15M.pt
-            export MODEL_NAME=stories15M
-            export MODEL_DIR=/tmp
-            for DTYPE in float16 float32; do
-              # if [ $(uname -s) == Darwin ]; then
-              #   export DTYPE=float16
-              # fi
-
-              python3 torchchat.py generate --dtype ${DTYPE} --device mps --checkpoint-path ${MODEL_PATH} --temperature 0
-
-              python3 torchchat.py generate --dtype ${DTYPE} --device mps --quant '{"embedding" : {"bitwidth": 8, "groupsize": 0}}' --checkpoint-path ${MODEL_PATH} --temperature 0
-
-              python3 torchchat.py generate --dtype ${DTYPE} --device mps --quant '{"embedding" : {"bitwidth": 8, "groupsize": 8}}' --checkpoint-path ${MODEL_PATH} --temperature 0
-
-              python3 torchchat.py generate --dtype ${DTYPE} --device mps --quant '{"linear:int8" : {"bitwidth": 8, "groupsize": 0}}' --checkpoint-path ${MODEL_PATH} --temperature 0
-
-              python3 torchchat.py generate --dtype ${DTYPE} --device mps --quant '{"linear:int8" : {"bitwidth": 8, "groupsize": 8}}' --checkpoint-path ${MODEL_PATH} --temperature 0
-
-              PYTORCH_ENABLE_MPS_FALLBACK=1 python3 torchchat.py generate --dtype ${DTYPE} --device mps --quant '{"linear:int4" : {"groupsize": 32}}' --checkpoint-path ${MODEL_PATH} --temperature 0
+    steps:
+      - name: Setup Python
+        uses: actions/setup-python@v2
+        with:
+          python-version: 3.10.11
+      - name: Run test
+        run: |
+          export PYTHON_VERSION="3.10"
+          set -x
+          # NS/MC: Remove previous installation of torch and torchao first
+          # as this script does not install anything into conda env but rather as system dep
+          pip3 uninstall -y torch || true
+          set -eou pipefail
+
+          pip3 uninstall -y torchao || true
+          set -eou pipefail
+
+          echo "::group::Print machine info"
+          uname -a
+          sysctl machdep.cpu.brand_string
+          sysctl machdep.cpu.core_count
+          echo "::endgroup::"
+
+          echo "::group::Install requirements"
+          # Install requirements
+          ./install/install_requirements.sh
+          ls -la
+          pwd
+          pip3 list
+          python3 -c 'import torch;print(f"torch: {torch.__version__, torch.version.git_version}")'
+          echo "::endgroup::"
+
+          echo "::group::Download checkpoints"
+          (
+            mkdir -p checkpoints/stories15M
+            pushd checkpoints/stories15M
+            curl -fsSL -O https://huggingface.co/karpathy/tinyllamas/resolve/main/stories15M.pt
+            curl -fsSL -O https://github.com/karpathy/llama2.c/raw/master/tokenizer.model
+            popd
+          )
+          echo "::endgroup::"
+
+          echo "::group::Run inference"
+          export MODEL_PATH=checkpoints/stories15M/stories15M.pt
+          export MODEL_NAME=stories15M
+          export MODEL_DIR=/tmp
+          for DTYPE in float16 float32; do
+            # if [ $(uname -s) == Darwin ]; then
+            #   export DTYPE=float16
+            # fi
+
+            python3 torchchat.py generate --dtype ${DTYPE} --device mps --checkpoint-path ${MODEL_PATH} --temperature 0
+
+            python3 torchchat.py generate --dtype ${DTYPE} --device mps --quant '{"embedding" : {"bitwidth": 8, "groupsize": 0}}' --checkpoint-path ${MODEL_PATH} --temperature 0
+
+            python3 torchchat.py generate --dtype ${DTYPE} --device mps --quant '{"embedding" : {"bitwidth": 8, "groupsize": 8}}' --checkpoint-path ${MODEL_PATH} --temperature 0
+
+            python3 torchchat.py generate --dtype ${DTYPE} --device mps --quant '{"linear:int8" : {"bitwidth": 8, "groupsize": 0}}' --checkpoint-path ${MODEL_PATH} --temperature 0
+
+            python3 torchchat.py generate --dtype ${DTYPE} --device mps --quant '{"linear:int8" : {"bitwidth": 8, "groupsize": 8}}' --checkpoint-path ${MODEL_PATH} --temperature 0
+
+            PYTORCH_ENABLE_MPS_FALLBACK=1 python3 torchchat.py generate --dtype ${DTYPE} --device mps --quant '{"linear:int4" : {"groupsize": 32}}' --checkpoint-path ${MODEL_PATH} --temperature 0
   compile-gguf:
     strategy:
       matrix: