[fix] baseline4 docker build move ex_engine into qwen3_6_scripts, remove COPY ex_engine from Dockerfile

qwen3_6_scripts/ex_engine/xllm_kernels/build/tmp_hgemm_blocktiling/.ninja_log

@@ -0,0 +1,4 @@
+# ninja log v5
+0	604	1786724827165020649	hgemm_blocktiling.cuda.o	3d7c5ae846fcf56e
+0	16465	1786726494208594547	hgemm_bind.o	808024d62d52e3c9
+16465	16725	1786726494464597555	hgemm_blocktiling.so	be64c512cf1529ab

qwen3_6_scripts/ex_engine/xllm_kernels/build/tmp_hgemm_blocktiling/build.ninja

@@ -0,0 +1,32 @@
+ninja_required_version = 1.3
+cxx = c++
+nvcc = /usr/local/corex/bin/clang++
+
+cflags = -DTORCH_EXTENSION_NAME=hgemm_blocktiling -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1011\" -I/home/dylan/0814/project_6/ex_engine/xllm_kernels/cuda/headers -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/TH -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/THC -isystem /usr/local/corex/include -isystem /usr/local/include/python3.10 -D_GLIBCXX_USE_CXX11_ABI=0 -fPIC -std=c++17 -O2 -std=c++17
+post_cflags = 
+cuda_cflags = -DTORCH_EXTENSION_NAME=hgemm_blocktiling -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1011\" -I/home/dylan/0814/project_6/ex_engine/xllm_kernels/cuda/headers -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/TH -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/THC -isystem /usr/local/corex/include -isystem /usr/local/include/python3.10 -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ -D__ILUVATAR__ -D__ILUVATAR_WORKAROUND__ -D__ILUVATAR_DIAG__ -cl-single-precision-constant -fPIC -mllvm --bonus-inst-threshold=0 -O2 --cuda-gpu-arch=ivcore10 --cuda-path=/usr/local/corex -std=c++17
+cuda_post_cflags = 
+cuda_dlink_post_cflags = 
+ldflags = -shared -L/usr/local/corex/lib64/python3/dist-packages/torch/lib -lc10 -lc10_cuda -ltorch_cpu -ltorch_cuda -ltorch -ltorch_python -L/usr/local/corex/lib64 -lcudart
+
+rule compile
+  command = $cxx -MMD -MF $out.d $cflags -c $in -o $out $post_cflags
+  depfile = $out.d
+  deps = gcc
+
+rule cuda_compile
+  command = $nvcc  $cuda_cflags -c $in -o $out $cuda_post_cflags
+
+
+
+rule link
+  command = $cxx $in $ldflags -o $out
+
+build hgemm_blocktiling.cuda.o: cuda_compile /home/dylan/0814/project_6/ex_engine/xllm_kernels/cuda/hgemm_blocktiling.cu
+build hgemm_bind.o: compile /home/dylan/0814/project_6/ex_engine/xllm_kernels/cuda/bindings/hgemm_bind.cpp
+
+
+
+build hgemm_blocktiling.so: link hgemm_blocktiling.cuda.o hgemm_bind.o
+
+default hgemm_blocktiling.so

qwen3_6_scripts/ex_engine/xllm_kernels/build/tmp_hgemm_warptiling/.ninja_log

@@ -0,0 +1,4 @@
+# ninja log v5
+0	651	1786770884612898246	hgemm_warptiling.cuda.o	a06bd2d038ca701e
+0	17136	1786770901105087181	hgemm_warp_bind.o	7a684f5698743bfe
+17136	17403	1786770901365090159	hgemm_warptiling.so	bbcad77d7d8433af

qwen3_6_scripts/ex_engine/xllm_kernels/build/tmp_hgemm_warptiling/build.ninja

@@ -0,0 +1,32 @@
+ninja_required_version = 1.3
+cxx = c++
+nvcc = /usr/local/corex/bin/clang++
+
+cflags = -DTORCH_EXTENSION_NAME=hgemm_warptiling -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1011\" -I/home/dylan/0814/project_6/ex_engine/xllm_kernels/cuda/headers -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/TH -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/THC -isystem /usr/local/corex/include -isystem /usr/local/include/python3.10 -D_GLIBCXX_USE_CXX11_ABI=0 -fPIC -std=c++17 -O2 -std=c++17
+post_cflags = 
+cuda_cflags = -DTORCH_EXTENSION_NAME=hgemm_warptiling -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1011\" -I/home/dylan/0814/project_6/ex_engine/xllm_kernels/cuda/headers -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/TH -isystem /usr/local/corex/lib64/python3/dist-packages/torch/include/THC -isystem /usr/local/corex/include -isystem /usr/local/include/python3.10 -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ -D__ILUVATAR__ -D__ILUVATAR_WORKAROUND__ -D__ILUVATAR_DIAG__ -cl-single-precision-constant -fPIC -mllvm --bonus-inst-threshold=0 -O2 --cuda-gpu-arch=ivcore10 --cuda-path=/usr/local/corex -std=c++17
+cuda_post_cflags = 
+cuda_dlink_post_cflags = 
+ldflags = -shared -L/usr/local/corex/lib64/python3/dist-packages/torch/lib -lc10 -lc10_cuda -ltorch_cpu -ltorch_cuda -ltorch -ltorch_python -L/usr/local/corex/lib64 -lcudart
+
+rule compile
+  command = $cxx -MMD -MF $out.d $cflags -c $in -o $out $post_cflags
+  depfile = $out.d
+  deps = gcc
+
+rule cuda_compile
+  command = $nvcc  $cuda_cflags -c $in -o $out $cuda_post_cflags
+
+
+
+rule link
+  command = $cxx $in $ldflags -o $out
+
+build hgemm_warptiling.cuda.o: cuda_compile /home/dylan/0814/project_6/ex_engine/xllm_kernels/cuda/hgemm_warptiling.cu
+build hgemm_warp_bind.o: compile /home/dylan/0814/project_6/ex_engine/xllm_kernels/cuda/bindings/hgemm_warp_bind.cpp
+
+
+
+build hgemm_warptiling.so: link hgemm_warptiling.cuda.o hgemm_warp_bind.o
+
+default hgemm_warptiling.so