[Bug] chatModule.resetChat() not always work well on iOS and Llama 2

[ruiba168]

🐛 Bug

chatModule.resetChat() is used to clear session info. If i call chatModule.resetChat() before every prefill call, previous prompt should not influent next round's out. But i found if i keep do prefill/decode/resetChat about 10-20 round later, sound resetChat() didn't reset it well, the decoded result will try to answer previous prompt and current prompt together

To Reproduce

Steps to reproduce the behavior:

1. Just use MLCChat iOS app to do test

Expected behavior

Environment

• Platform (e.g. WebGPU/Vulkan/IOS/Android/CUDA):iOS
• Operating system (e.g. Ubuntu/Windows/MacOS/...):MacOS
• Device (e.g. iPhone 12 Pro, PC+RTX 3090, ...):MyMac(Designed for iPad)
• How you installed MLC-LLM (conda, source):conda
• How you installed TVM-Unity (pip, source): pip
• Python version (e.g. 3.10): 3.11
• 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):
• Any other relevant information:
• USE_NVTX: OFF USE_GTEST: AUTO SUMMARIZE: OFF USE_IOS_RPC: OFF USE_MSC: OFF USE_ETHOSU: 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: 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 --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: 5e06dab1a3bc691d71230d56d3bb3bb9df1220ce 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: 2023-11-12 15:22:32 -0500 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: 15.0.7 USE_OPENCL: OFF COMPILER_RT_PATH: 3rdparty/compiler-rt RANG_PATH: 3rdparty/rang/include USE_SPIRV_KHR_INTEGER_DOT_PRODUCT: OFF USE_OPENMP: OFF 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++

Additional context

[tqchen]

Thank you for reporting this. do you mind to try and see if we can also reproduce this behavior through python API? This would help us better diagnose potential issues(as python api and iOS uses the same backend impl).

[davidpissarra]

Hi @ruiba168, any updates? Can you share some concrete steps on how to reproduce the issue you are facing, so I can follow up on this? Thanks a lot!