Operating system (e.g. Ubuntu/Windows/MacOS/...): macOS
Device (e.g. iPhone 12 Pro, PC+RTX 3090, ...): M2 Mac Studio Max
How you installed MLC-LLM (conda, source): git clone
How you installed TVM-Unity (pip, source): git clone
Python version (e.g. 3.10): 3.10
GPU driver version (if applicable):
CUDA/cuDNN version (if applicable):
TVM Unity Hash Tag (python -c "import tvm; print('\n'.join(f'{k}: {v}' for k, v in tvm.support.libinfo().items()))", applicable if you compile models):
USE_GTEST: AUTO
SUMMARIZE: OFF
USE_IOS_RPC: OFF
USE_MSC: OFF
USE_ETHOSU: OFF
CUDA_VERSION: NOT-FOUND
USE_LIBBACKTRACE: AUTO
DLPACK_PATH: 3rdparty/dlpack/include
USE_TENSORRT_CODEGEN: OFF
USE_THRUST: OFF
USE_TARGET_ONNX: OFF
USE_AOT_EXECUTOR: ON
BUILD_DUMMY_LIBTVM: OFF
USE_CUDNN: OFF
USE_TENSORRT_RUNTIME: OFF
USE_ARM_COMPUTE_LIB_GRAPH_EXECUTOR: OFF
USE_CCACHE: AUTO
USE_ARM_COMPUTE_LIB: OFF
USE_CPP_RTVM: OFF
USE_OPENCL_GTEST: /path/to/opencl/gtest
USE_MKL: OFF
USE_PT_TVMDSOOP: OFF
MLIR_VERSION: NOT-FOUND
USE_CLML: OFF
USE_STACKVM_RUNTIME: OFF
USE_GRAPH_EXECUTOR_CUDA_GRAPH: OFF
ROCM_PATH: /opt/rocm
USE_DNNL: OFF
USE_VITIS_AI: OFF
USE_MLIR: OFF
USE_RCCL: OFF
USE_LLVM: llvm-config --ignore-libllvm --link-static
USE_VERILATOR: OFF
USE_TF_TVMDSOOP: OFF
USE_THREADS: ON
USE_MSVC_MT: OFF
BACKTRACE_ON_SEGFAULT: OFF
USE_GRAPH_EXECUTOR: ON
USE_NCCL: OFF
USE_ROCBLAS: OFF
GIT_COMMIT_HASH: 761cae9c7195be901beb09450db0b8a447f7ccba
USE_VULKAN: OFF
USE_RUST_EXT: OFF
USE_CUTLASS: OFF
USE_CPP_RPC: OFF
USE_HEXAGON: OFF
USE_CUSTOM_LOGGING: OFF
USE_UMA: OFF
USE_FALLBACK_STL_MAP: OFF
USE_SORT: ON
USE_RTTI: ON
GIT_COMMIT_TIME: 2023-09-09 13:47:14 -0400
USE_HEXAGON_SDK: /path/to/sdk
USE_BLAS: none
USE_ETHOSN: OFF
USE_LIBTORCH: OFF
USE_RANDOM: ON
USE_CUDA: OFF
USE_COREML: OFF
USE_AMX: OFF
BUILD_STATIC_RUNTIME: OFF
USE_CMSISNN: OFF
USE_KHRONOS_SPIRV: OFF
USE_CLML_GRAPH_EXECUTOR: OFF
USE_TFLITE: OFF
USE_HEXAGON_GTEST: /path/to/hexagon/gtest
PICOJSON_PATH: 3rdparty/picojson
USE_OPENCL_ENABLE_HOST_PTR: OFF
INSTALL_DEV: OFF
USE_PROFILER: ON
USE_NNPACK: OFF
LLVM_VERSION: 16.0.6
USE_OPENCL: OFF
COMPILER_RT_PATH: 3rdparty/compiler-rt
RANG_PATH: 3rdparty/rang/include
USE_SPIRV_KHR_INTEGER_DOT_PRODUCT: OFF
USE_OPENMP: none
USE_BNNS: OFF
USE_CUBLAS: OFF
USE_METAL: ON
USE_MICRO_STANDALONE_RUNTIME: OFF
USE_HEXAGON_EXTERNAL_LIBS: OFF
USE_ALTERNATIVE_LINKER: AUTO
USE_BYODT_POSIT: OFF
USE_HEXAGON_RPC: OFF
USE_MICRO: OFF
DMLC_PATH: 3rdparty/dmlc-core/include
INDEX_DEFAULT_I64: ON
USE_RELAY_DEBUG: OFF
USE_RPC: ON
USE_TENSORFLOW_PATH: none
TVM_CLML_VERSION:
USE_MIOPEN: OFF
USE_ROCM: OFF
USE_PAPI: OFF
USE_CURAND: OFF
TVM_CXX_COMPILER_PATH: /Library/Developer/CommandLineTools/usr/bin/c++
HIDE_PRIVATE_SYMBOLS: ON
jeethu
commented
1 year ago
🐛 Bug
Compiling models on macOS (I've tried this on x86 and Apple Silicon Macs) fails with syntax errors in the generated metal source.
To Reproduce
Steps to reproduce the behavior:
python build.py --model path/to/Llama-2-7b-chat-hf --quantization q4f16_1 --target iphonepython3 build.py --model ~/LLM_models/openchat_v3.2_super --quantization q4f16_1 --target metalExpected behavior
Environment
conda, source): git clonepip, source): git clonepython -c "import tvm; print('\n'.join(f'{k}: {v}' for k, v in tvm.support.libinfo().items()))", applicable if you compile models):Additional context
893 is a related issue.