yuvalkirstain's picture
add clip
17db41a
Raw
History Blame Contribute Delete
2.39 kB
CC = nvcc
CFLAGS =
fPIC_flag = -Xcompiler=-fPIC
obj_files = conv2d_bias_1541.obj conv2d_bias_1542.obj groupnorm_swish_1543.obj conv2d_bias_1544.obj conv2d_bias_add_identity_1638.obj groupnorm_1548.obj gemm_rcr_bias_permute_1551.obj split_1553.obj reshape_1554.obj mem_eff_attention_1557.obj gemm_rcr_bias_add_1559.obj fused_elementwise_1652.obj upsampling2d_1583.obj conv2d_bias_1584.obj groupnorm_swish_1585.obj conv2d_bias_add_identity_1643.obj upsampling2d_1600.obj conv2d_bias_1601.obj groupnorm_swish_1602.obj conv2d_bias_1603.obj groupnorm_swish_1604.obj conv2d_bias_1606.obj conv2d_bias_add_identity_1646.obj conv2d_bias_1609.obj upsampling2d_1618.obj conv2d_bias_1619.obj groupnorm_swish_1620.obj conv2d_bias_1621.obj groupnorm_swish_1622.obj conv2d_bias_1624.obj conv2d_bias_add_identity_1649.obj conv2d_bias_1627.obj conv2d_bias_1637.obj constants.obj model_container_base.obj debug_utility.obj model_container.obj model_interface.obj utility.obj
%.obj : %.cu
nvcc -DCUTLASS_ENABLE_TENSOR_CORE_MMA=1 -DCUTLASS_USE_TANH_FOR_SIGMOID=1 -w -gencode=arch=compute_80,code=[sm_80,compute_80] -Xcompiler=-fPIC -Xcompiler=-Wconversion -Xcompiler=-fno-strict-aliasing -Xcompiler -fvisibility=hidden -O3 -std=c++17 --expt-relaxed-constexpr --use_fast_math -I/home/olab/kirstain/anaconda3/envs/ait/lib/python3.10/site-packages/aitemplate/3rdparty/cutlass/include -I/home/olab/kirstain/anaconda3/envs/ait/lib/python3.10/site-packages/aitemplate/3rdparty/cutlass/tools/util/include -I/home/olab/kirstain/anaconda3/envs/ait/lib/python3.10/site-packages/aitemplate/3rdparty/cutlass/examples/35_gemm_softmax -I/home/olab/kirstain/anaconda3/envs/ait/lib/python3.10/site-packages/aitemplate/3rdparty/cutlass/examples/41_fused_multi_head_attention -I/home/olab/kirstain/anaconda3/envs/ait/lib/python3.10/site-packages/aitemplate/3rdparty/cutlass/examples/45_dual_gemm -I/home/olab/kirstain/anaconda3/envs/ait/lib/python3.10/site-packages/aitemplate/3rdparty/cutlass/../../backend/cuda/attention/src/./ -I/home/olab/kirstain/anaconda3/envs/ait/lib/python3.10/site-packages/aitemplate/3rdparty/cutlass/../../backend/cuda/attention/src/fmha -DNDEBUG -c -o $@ $<
%.obj : %.bin
ld -r -b binary -o $@ $<
.PHONY: all clean clean_constants
all: test.so
test.so: $(obj_files)
$(CC) -shared $(fPIC_flag) $(CFLAGS) -o $@ $(obj_files)
clean:
rm -f *.obj test.so
clean_constants:
rm -f constants.bin