wiki-documents
wiki-documents copied to clipboard
Seeed-Studio
Error converting ONNX following the example
I've followed the example in the documentation, trying to convert the provided yolo11n.onnx to cvimodel for seeedstudio recamera
root@60b080763dad:/workspace# model_deploy \ --mlir yolo11n.mlir \ --quant_input \ --quantize F16 \
--customization_format RGB_PACKED \
--processor cv181x \
--test_input car.jpg \
--test_reference yolo11n_top_outputs.npz \
--fuse_preprocess \
--tolerance 0.99,0.9 \
--model yolo11n_1684x_f16.cvimodel
OpenBLAS WARNING - could not determine the L2 cache size on this system, assuming 256k
OpenBLAS WARNING - could not determine the L2 cache size on this system, assuming 256k
2025/02/19 15:50:30 - INFO : TPU-MLIR v1.15.1-20250208
2025/02/19 15:50:30 - INFO :
load_config Preprocess args :
resize_dims : [640, 640]
keep_aspect_ratio : True
keep_ratio_mode : letterbox
pad_value : 0
pad_type : center
input_dims : [640, 640]
--------------------------
mean : [0.0, 0.0, 0.0]
scale : [0.0039216, 0.0039216, 0.0039216]
--------------------------
pixel_format : rgb
channel_format : nchw
2025/02/19 15:50:30 - INFO : Add preprocess, set the following params:
2025/02/19 15:50:30 - INFO :
_____________________________________________________
| preprocess: |
| (x - mean) * scale |
'-------------------------------------------------------'
config Preprocess args :
resize_dims : [640, 640]
keep_aspect_ratio : True
keep_ratio_mode : letterbox
pad_value : 0
pad_type : center
--------------------------
mean : [0.0, 0.0, 0.0]
scale : [1.0, 1.0, 1.0]
--------------------------
pixel_format : rgb
channel_format : nhwc
[Running]: tpuc-opt yolo11n.mlir --processor-assign="chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize --fuse-preprocess="mode=F16 customization_format=RGB_PACKED align=False" --convert-top-to-tpu=" asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold -o yolo11n_cv181x_f16_tpu.mlir
ConvertUpsampleOp : top.Upsample succeed!
ConvertUpsampleOp : top.Upsample succeed!
Entering FusePreprocessPass.
raw format not support current type
UNREACHABLE executed at /__w/tpu-mlir/tpu-mlir/lib/Conversion/TopToTpu/CV18xx/Preprocess.cpp:108!
PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace.
Stack dump:
0. Program arguments: tpuc-opt yolo11n.mlir --init "--processor-assign=chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize "--fuse-preprocess=mode=F16 customization_format=RGB_PACKED align=False" "--convert-top-to-tpu= asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold --deinit --mlir-print-debuginfo -o yolo11n_cv181x_f16_tpu.mlir
#0 0x000055b3dfaaf467 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884467)
#1 0x000055b3dfaad18e (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x88218e)
#2 0x000055b3dfaafdea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884dea)
#3 0x00007afc29347520 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42520)
#4 0x00007afc2939b9fc pthread_kill (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x969fc)
#5 0x00007afc29347476 gsignal (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42476)
#6 0x00007afc2932d7f3 abort (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x287f3)
#7 0x000055b3dfaacfb1 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x881fb1)
#8 0x000055b3dff00850 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xcd5850)
#9 0x000055b3dfd5d656 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb32656)
#10 0x000055b3dfd5cf34 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb31f34)
#11 0x000055b3e10ee337 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ec3337)
#12 0x000055b3e10eac4f (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ebfc4f)
#13 0x000055b3e10b3d2c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e88d2c)
#14 0x000055b3e10b0b7c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e85b7c)
#15 0x000055b3dfc0b394 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x9e0394)
#16 0x000055b3e1117624 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eec624)
#17 0x000055b3e1117c51 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eecc51)
#18 0x000055b3e111a0f8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eef0f8)
#19 0x000055b3dfaa0b1b (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x875b1b)
#20 0x000055b3dfa9fee4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x874ee4)
#21 0x000055b3e1334ac8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x2109ac8)
#22 0x000055b3dfa9a1ea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f1ea)
#23 0x000055b3dfa9a6b4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f6b4)
#24 0x000055b3dfa990fa (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86e0fa)
#25 0x00007afc2932ed90 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29d90)
#26 0x00007afc2932ee40 __libc_start_main (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29e40)
#27 0x000055b3dfa98505 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86d505)
Aborted
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 556, in <module>
lowering_patterns = tool.lowering()
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 200, in lowering
patterns = mlir_lowering(
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 559, in mlir_lowering
_os_system(cmd, mute=mute,log_level=log_level)
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 420, in _os_system
raise RuntimeError("[!Error]: {}".format(cmd_str))
RuntimeError: [!Error]: tpuc-opt yolo11n.mlir --processor-assign="chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize --fuse-preprocess="mode=F16 customization_format=RGB_PACKED align=False" --convert-top-to-tpu=" asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold -o yolo11n_cv181x_f16_tpu.mlir
root@60b080763dad:/workspace#
👋 @webfrank
Thank you for raising an issue. We will investigate into the matter and get back to you as soon as possible. Please make sure you have given us as much context as possible.
![github-actions[bot] avatar](https://avatars.githubusercontent.com/in/15368?v=4 "Published by a pub.dev verified publisher"){kind=small}
And your command seems to lack some parameters.
I've followed the example in the documentation, trying to convert the provided yolo11n.onnx to cvimodel for seeedstudio recamera
root@60b080763dad:/workspace# model_deploy \ --mlir yolo11n.mlir \ --quant_input \ --quantize F16 \ --customization_format RGB_PACKED \ --processor cv181x \ --test_input car.jpg \ --test_reference yolo11n_top_outputs.npz \ --fuse_preprocess \ --tolerance 0.99,0.9 \ --model yolo11n_1684x_f16.cvimodel OpenBLAS WARNING - could not determine the L2 cache size on this system, assuming 256k OpenBLAS WARNING - could not determine the L2 cache size on this system, assuming 256k 2025/02/19 15:50:30 - INFO : TPU-MLIR v1.15.1-20250208 2025/02/19 15:50:30 - INFO : load_config Preprocess args : resize_dims : [640, 640] keep_aspect_ratio : True keep_ratio_mode : letterbox pad_value : 0 pad_type : center input_dims : [640, 640] -------------------------- mean : [0.0, 0.0, 0.0] scale : [0.0039216, 0.0039216, 0.0039216] -------------------------- pixel_format : rgb channel_format : nchw 2025/02/19 15:50:30 - INFO : Add preprocess, set the following params: 2025/02/19 15:50:30 - INFO : _____________________________________________________ | preprocess: | | (x - mean) * scale | '-------------------------------------------------------' config Preprocess args : resize_dims : [640, 640] keep_aspect_ratio : True keep_ratio_mode : letterbox pad_value : 0 pad_type : center -------------------------- mean : [0.0, 0.0, 0.0] scale : [1.0, 1.0, 1.0] -------------------------- pixel_format : rgb channel_format : nhwc [Running]: tpuc-opt yolo11n.mlir --processor-assign="chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize --fuse-preprocess="mode=F16 customization_format=RGB_PACKED align=False" --convert-top-to-tpu=" asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold -o yolo11n_cv181x_f16_tpu.mlir ConvertUpsampleOp : top.Upsample succeed! ConvertUpsampleOp : top.Upsample succeed! Entering FusePreprocessPass. raw format not support current type UNREACHABLE executed at /__w/tpu-mlir/tpu-mlir/lib/Conversion/TopToTpu/CV18xx/Preprocess.cpp:108! PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace. Stack dump: 0. Program arguments: tpuc-opt yolo11n.mlir --init "--processor-assign=chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize "--fuse-preprocess=mode=F16 customization_format=RGB_PACKED align=False" "--convert-top-to-tpu= asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold --deinit --mlir-print-debuginfo -o yolo11n_cv181x_f16_tpu.mlir #0 0x000055b3dfaaf467 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884467) #1 0x000055b3dfaad18e (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x88218e) #2 0x000055b3dfaafdea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884dea) #3 0x00007afc29347520 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42520) #4 0x00007afc2939b9fc pthread_kill (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x969fc) #5 0x00007afc29347476 gsignal (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42476) #6 0x00007afc2932d7f3 abort (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x287f3) #7 0x000055b3dfaacfb1 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x881fb1) #8 0x000055b3dff00850 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xcd5850) #9 0x000055b3dfd5d656 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb32656) #10 0x000055b3dfd5cf34 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb31f34) #11 0x000055b3e10ee337 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ec3337) #12 0x000055b3e10eac4f (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ebfc4f) #13 0x000055b3e10b3d2c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e88d2c) #14 0x000055b3e10b0b7c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e85b7c) #15 0x000055b3dfc0b394 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x9e0394) #16 0x000055b3e1117624 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eec624) #17 0x000055b3e1117c51 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eecc51) #18 0x000055b3e111a0f8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eef0f8) #19 0x000055b3dfaa0b1b (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x875b1b) #20 0x000055b3dfa9fee4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x874ee4) #21 0x000055b3e1334ac8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x2109ac8) #22 0x000055b3dfa9a1ea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f1ea) #23 0x000055b3dfa9a6b4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f6b4) #24 0x000055b3dfa990fa (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86e0fa) #25 0x00007afc2932ed90 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29d90) #26 0x00007afc2932ee40 __libc_start_main (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29e40) #27 0x000055b3dfa98505 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86d505) Aborted Traceback (most recent call last): File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 556, in <module> lowering_patterns = tool.lowering() File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 200, in lowering patterns = mlir_lowering( File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 559, in mlir_lowering _os_system(cmd, mute=mute,log_level=log_level) File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 420, in _os_system raise RuntimeError("[!Error]: {}".format(cmd_str)) RuntimeError: [!Error]: tpuc-opt yolo11n.mlir --processor-assign="chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize --fuse-preprocess="mode=F16 customization_format=RGB_PACKED align=False" --convert-top-to-tpu=" asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold -o yolo11n_cv181x_f16_tpu.mlir root@60b080763dad:/workspace#
Your command seems to lack some parameters.
Model is in the required version
Full commands are:
pip install tpu_mlir[onnx]
model_transform \
--model_name alpr \
--model_def anpr-demo-model.onnx \
--input_shapes "[[1,3,640,640]]" \
--mean "0.0,0.0,0.0" \
--scale "0.0039216,0.0039216,0.0039216" \
--keep_aspect_ratio \
--pixel_format rgb \
--output_names "/model.23/cv2.0/cv2.0.2/Conv_output_0,/model.23/cv3.0/cv3.0.2/Conv_output_0,/model.23/cv2.1/cv2.1.2/Conv_output_0,/model.23/cv3.1/cv3.1.2/Conv_output_0,/model.23/cv2.2/cv2.2.2/Conv_output_0,/model.23/cv3.2/cv3.2.2/Conv_output_0" \
--test_input car.png \
--test_result alpr_top_outputs.npz \
--mlir alpr.mlir
The final output is
min_similiarity = (0.9999995231628418, 0.9999982078713782, 114.61116790771484)
Target alpr_top_outputs.npz
Reference alpr_ref_outputs.npz
npz compare PASSED.
compare /model.23/cv3.2/cv3.2.2/Conv_output_0_Conv: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████| 230/230 [00:22<00:00, 10.39it/s]
[Success]: npz_tool.py compare alpr_top_outputs.npz alpr_ref_outputs.npz --tolerance 0.99,0.99 --except - -vv
model_deploy \
--mlir alpr.mlir \
--quant_input \
--customization_format RGB_PACKED \
--quantize F16 \
--processor cv181x \
--test_input car.png \
--test_reference alpr_top_outputs.npz \
--fuse_preprocess \
--tolerance 0.99,0.99 \
--model alpr.cvimodel
and the error is the same
ConvertUpsampleOp : top.Upsample succeed!
Entering FusePreprocessPass.
raw format not support current type
UNREACHABLE executed at /__w/tpu-mlir/tpu-mlir/lib/Conversion/TopToTpu/CV18xx/Preprocess.cpp:108!
PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace.
Stack dump:
0. Program arguments: tpuc-opt alpr.mlir --init "--processor-assign=chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize "--fuse-preprocess=mode=F16 customization_format=RGB_PACKED align=False" "--convert-top-to-tpu= asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold --deinit --mlir-print-debuginfo -o alpr_cv181x_f16_tpu.mlir
#0 0x00005f0ed2b50467 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884467)
#1 0x00005f0ed2b4e18e (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x88218e)
#2 0x00005f0ed2b50dea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884dea)
#3 0x0000781cf1788520 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42520)
#4 0x0000781cf17dc9fc pthread_kill (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x969fc)
#5 0x0000781cf1788476 gsignal (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42476)
#6 0x0000781cf176e7f3 abort (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x287f3)
#7 0x00005f0ed2b4dfb1 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x881fb1)
#8 0x00005f0ed2fa1850 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xcd5850)
#9 0x00005f0ed2dfe656 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb32656)
#10 0x00005f0ed2dfdf34 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb31f34)
#11 0x00005f0ed418f337 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ec3337)
#12 0x00005f0ed418bc4f (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ebfc4f)
#13 0x00005f0ed4154d2c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e88d2c)
#14 0x00005f0ed4151b7c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e85b7c)
#15 0x00005f0ed2cac394 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x9e0394)
#16 0x00005f0ed41b8624 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eec624)
#17 0x00005f0ed41b8c51 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eecc51)
#18 0x00005f0ed41bb0f8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eef0f8)
#19 0x00005f0ed2b41b1b (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x875b1b)
#20 0x00005f0ed2b40ee4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x874ee4)
#21 0x00005f0ed43d5ac8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x2109ac8)
#22 0x00005f0ed2b3b1ea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f1ea)
#23 0x00005f0ed2b3b6b4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f6b4)
#24 0x00005f0ed2b3a0fa (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86e0fa)
#25 0x0000781cf176fd90 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29d90)
#26 0x0000781cf176fe40 __libc_start_main (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29e40)
#27 0x00005f0ed2b39505 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86d505)
Aborted
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 556, in <module>
lowering_patterns = tool.lowering()
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 200, in lowering
patterns = mlir_lowering(
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 559, in mlir_lowering
_os_system(cmd, mute=mute,log_level=log_level)
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 420, in _os_system
raise RuntimeError("[!Error]: {}".format(cmd_str))
RuntimeError: [!Error]: tpuc-opt alpr.mlir --processor-assign="chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize --fuse-preprocess="mode=F16 customization_format=RGB_PACKED align=False" --convert-top-to-tpu=" asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold -o alpr_cv181x_f16_tpu.mlir
I tried also with the suggested yolo11n.onnx but results are the same. Tested with docker image 3.1, 3.2 and 3.3
Model is in the required version
Full commands are:
pip install tpu_mlir[onnx]model_transform \ --model_name alpr \ --model_def anpr-demo-model.onnx \ --input_shapes "[[1,3,640,640]]" \ --mean "0.0,0.0,0.0" \ --scale "0.0039216,0.0039216,0.0039216" \ --keep_aspect_ratio \ --pixel_format rgb \ --output_names "/model.23/cv2.0/cv2.0.2/Conv_output_0,/model.23/cv3.0/cv3.0.2/Conv_output_0,/model.23/cv2.1/cv2.1.2/Conv_output_0,/model.23/cv3.1/cv3.1.2/Conv_output_0,/model.23/cv2.2/cv2.2.2/Conv_output_0,/model.23/cv3.2/cv3.2.2/Conv_output_0" \ --test_input car.png \ --test_result alpr_top_outputs.npz \ --mlir alpr.mlirThe final output is
min_similiarity = (0.9999995231628418, 0.9999982078713782, 114.61116790771484) Target alpr_top_outputs.npz Reference alpr_ref_outputs.npz npz compare PASSED. compare /model.23/cv3.2/cv3.2.2/Conv_output_0_Conv: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████| 230/230 [00:22<00:00, 10.39it/s] [Success]: npz_tool.py compare alpr_top_outputs.npz alpr_ref_outputs.npz --tolerance 0.99,0.99 --except - -vvmodel_deploy \ --mlir alpr.mlir \ --quant_input \ --customization_format RGB_PACKED \ --quantize F16 \ --processor cv181x \ --test_input car.png \ --test_reference alpr_top_outputs.npz \ --fuse_preprocess \ --tolerance 0.99,0.99 \ --model alpr.cvimodeland the error is the same
ConvertUpsampleOp : top.Upsample succeed! Entering FusePreprocessPass. raw format not support current type UNREACHABLE executed at /__w/tpu-mlir/tpu-mlir/lib/Conversion/TopToTpu/CV18xx/Preprocess.cpp:108! PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace. Stack dump: 0. Program arguments: tpuc-opt alpr.mlir --init "--processor-assign=chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize "--fuse-preprocess=mode=F16 customization_format=RGB_PACKED align=False" "--convert-top-to-tpu= asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold --deinit --mlir-print-debuginfo -o alpr_cv181x_f16_tpu.mlir #0 0x00005f0ed2b50467 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884467) #1 0x00005f0ed2b4e18e (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x88218e) #2 0x00005f0ed2b50dea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884dea) #3 0x0000781cf1788520 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42520) #4 0x0000781cf17dc9fc pthread_kill (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x969fc) #5 0x0000781cf1788476 gsignal (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42476) #6 0x0000781cf176e7f3 abort (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x287f3) #7 0x00005f0ed2b4dfb1 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x881fb1) #8 0x00005f0ed2fa1850 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xcd5850) #9 0x00005f0ed2dfe656 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb32656) #10 0x00005f0ed2dfdf34 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb31f34) #11 0x00005f0ed418f337 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ec3337) #12 0x00005f0ed418bc4f (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ebfc4f) #13 0x00005f0ed4154d2c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e88d2c) #14 0x00005f0ed4151b7c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e85b7c) #15 0x00005f0ed2cac394 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x9e0394) #16 0x00005f0ed41b8624 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eec624) #17 0x00005f0ed41b8c51 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eecc51) #18 0x00005f0ed41bb0f8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eef0f8) #19 0x00005f0ed2b41b1b (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x875b1b) #20 0x00005f0ed2b40ee4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x874ee4) #21 0x00005f0ed43d5ac8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x2109ac8) #22 0x00005f0ed2b3b1ea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f1ea) #23 0x00005f0ed2b3b6b4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f6b4) #24 0x00005f0ed2b3a0fa (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86e0fa) #25 0x0000781cf176fd90 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29d90) #26 0x0000781cf176fe40 __libc_start_main (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29e40) #27 0x00005f0ed2b39505 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86d505) Aborted Traceback (most recent call last): File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 556, in <module> lowering_patterns = tool.lowering() File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 200, in lowering patterns = mlir_lowering( File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 559, in mlir_lowering _os_system(cmd, mute=mute,log_level=log_level) File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 420, in _os_system raise RuntimeError("[!Error]: {}".format(cmd_str)) RuntimeError: [!Error]: tpuc-opt alpr.mlir --processor-assign="chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize --fuse-preprocess="mode=F16 customization_format=RGB_PACKED align=False" --convert-top-to-tpu=" asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold -o alpr_cv181x_f16_tpu.mlirI tried also with the suggested yolo11n.onnx but results are the same. Tested with docker image 3.1, 3.2 and 3.3
hi, frank. Could you shared your ONNX file with me? i will test it.
Hi, I will upload it here, no problem, but the error is the same using the linked yolov11n.onnx
https://fastupload.io/71369425ce690b9a
Hi, any news on this? Other users got the same issue. I bought the cam to be able to use a custom trained model, not the preloaded one. Without this it is just a cute paperweight. How did you converted the standard yolo11n.onnx to cvimodel? The described procedure does not work.
Hi, any news on this? Other users got the same issue. I bought the cam to be able to use a custom trained model, not the preloaded one. Without this it is just a cute paperweight. How did you converted the standard yolo11n.onnx to cvimodel? The described procedure does not work.
Hi, frank. I have convert your onnx to F16 cvimodel successfully. you also failed using the ONNX provided by Wiki, so I think it might be due to --test_input parameter. If it still doesn't work, you can try removing the --test_input parameter.
No go, removed --test_input
[Success]: tpuc-opt yolo11n_origin.mlir --shape-infer --canonicalize --extra-optimize -o yolo11n.mlir
2025/03/03 02:06:29 - INFO : Mlir file generated:yolo11n.mlir
root@25f2c35c9495:/workspace# model_deploy \
--mlir yolo11n.mlir \
--quant_input \
--customization_format RGB_PACKED \
--quantize F16 \
--processor cv181x \
--fuse_preprocess \
--tolerance 0.99,0.99 \
--model yolo11n.cvimodel
OpenBLAS WARNING - could not determine the L2 cache size on this system, assuming 256k
OpenBLAS WARNING - could not determine the L2 cache size on this system, assuming 256k
2025/03/03 02:07:11 - INFO : TPU-MLIR v1.15.2-20250224
[Running]: tpuc-opt yolo11n.mlir --processor-assign="chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize --fuse-preprocess="mode=F16 customization_format=RGB_PACKED align=False" --convert-top-to-tpu=" asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold -o yolo11n_cv181x_f16_tpu.mlir
ConvertUpsampleOp : top.Upsample succeed!
ConvertUpsampleOp : top.Upsample succeed!
Entering FusePreprocessPass.
raw format not support current type
UNREACHABLE executed at /__w/tpu-mlir/tpu-mlir/lib/Conversion/TopToTpu/CV18xx/Preprocess.cpp:108!
PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace.
Stack dump:
0. Program arguments: tpuc-opt yolo11n.mlir --init "--processor-assign=chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize "--fuse-preprocess=mode=F16 customization_format=RGB_PACKED align=False" "--convert-top-to-tpu= asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold --deinit --mlir-print-debuginfo -o yolo11n_cv181x_f16_tpu.mlir
#0 0x0000559f0802e467 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884467)
#1 0x0000559f0802c18e (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x88218e)
#2 0x0000559f0802edea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884dea)
#3 0x000076dcca1f0520 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42520)
#4 0x000076dcca2449fc pthread_kill (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x969fc)
#5 0x000076dcca1f0476 gsignal (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42476)
#6 0x000076dcca1d67f3 abort (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x287f3)
#7 0x0000559f0802bfb1 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x881fb1)
#8 0x0000559f0847f850 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xcd5850)
#9 0x0000559f082dc656 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb32656)
#10 0x0000559f082dbf34 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb31f34)
#11 0x0000559f0966d337 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ec3337)
#12 0x0000559f09669c4f (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ebfc4f)
#13 0x0000559f09632d2c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e88d2c)
#14 0x0000559f0962fb7c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e85b7c)
#15 0x0000559f0818a394 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x9e0394)
#16 0x0000559f09696624 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eec624)
#17 0x0000559f09696c51 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eecc51)
#18 0x0000559f096990f8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eef0f8)
#19 0x0000559f0801fb1b (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x875b1b)
#20 0x0000559f0801eee4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x874ee4)
#21 0x0000559f098b3ac8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x2109ac8)
#22 0x0000559f080191ea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f1ea)
#23 0x0000559f080196b4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f6b4)
#24 0x0000559f080180fa (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86e0fa)
#25 0x000076dcca1d7d90 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29d90)
#26 0x000076dcca1d7e40 __libc_start_main (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29e40)
#27 0x0000559f08017505 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86d505)
Aborted
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 556, in <module>
lowering_patterns = tool.lowering()
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 200, in lowering
patterns = mlir_lowering(
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 559, in mlir_lowering
_os_system(cmd, mute=mute,log_level=log_level)
File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 420, in _os_system
raise RuntimeError("[!Error]: {}".format(cmd_str))
RuntimeError: [!Error]: tpuc-opt yolo11n.mlir --processor-assign="chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize --fuse-preprocess="mode=F16 customization_format=RGB_PACKED align=False" --convert-top-to-tpu=" asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold -o yolo11n_cv181x_f16_tpu.mlir
root@25f2c35c9495:/workspace#
No go, removed --test_input
[Success]: tpuc-opt yolo11n_origin.mlir --shape-infer --canonicalize --extra-optimize -o yolo11n.mlir 2025/03/03 02:06:29 - INFO : Mlir file generated:yolo11n.mlir root@25f2c35c9495:/workspace# model_deploy \ --mlir yolo11n.mlir \ --quant_input \ --customization_format RGB_PACKED \ --quantize F16 \ --processor cv181x \ --fuse_preprocess \ --tolerance 0.99,0.99 \ --model yolo11n.cvimodel OpenBLAS WARNING - could not determine the L2 cache size on this system, assuming 256k OpenBLAS WARNING - could not determine the L2 cache size on this system, assuming 256k 2025/03/03 02:07:11 - INFO : TPU-MLIR v1.15.2-20250224 [Running]: tpuc-opt yolo11n.mlir --processor-assign="chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize --fuse-preprocess="mode=F16 customization_format=RGB_PACKED align=False" --convert-top-to-tpu=" asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold -o yolo11n_cv181x_f16_tpu.mlir ConvertUpsampleOp : top.Upsample succeed! ConvertUpsampleOp : top.Upsample succeed! Entering FusePreprocessPass. raw format not support current type UNREACHABLE executed at /__w/tpu-mlir/tpu-mlir/lib/Conversion/TopToTpu/CV18xx/Preprocess.cpp:108! PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace. Stack dump: 0. Program arguments: tpuc-opt yolo11n.mlir --init "--processor-assign=chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize "--fuse-preprocess=mode=F16 customization_format=RGB_PACKED align=False" "--convert-top-to-tpu= asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold --deinit --mlir-print-debuginfo -o yolo11n_cv181x_f16_tpu.mlir #0 0x0000559f0802e467 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884467) #1 0x0000559f0802c18e (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x88218e) #2 0x0000559f0802edea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x884dea) #3 0x000076dcca1f0520 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42520) #4 0x000076dcca2449fc pthread_kill (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x969fc) #5 0x000076dcca1f0476 gsignal (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x42476) #6 0x000076dcca1d67f3 abort (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x287f3) #7 0x0000559f0802bfb1 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x881fb1) #8 0x0000559f0847f850 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xcd5850) #9 0x0000559f082dc656 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb32656) #10 0x0000559f082dbf34 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0xb31f34) #11 0x0000559f0966d337 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ec3337) #12 0x0000559f09669c4f (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1ebfc4f) #13 0x0000559f09632d2c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e88d2c) #14 0x0000559f0962fb7c (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1e85b7c) #15 0x0000559f0818a394 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x9e0394) #16 0x0000559f09696624 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eec624) #17 0x0000559f09696c51 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eecc51) #18 0x0000559f096990f8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x1eef0f8) #19 0x0000559f0801fb1b (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x875b1b) #20 0x0000559f0801eee4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x874ee4) #21 0x0000559f098b3ac8 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x2109ac8) #22 0x0000559f080191ea (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f1ea) #23 0x0000559f080196b4 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86f6b4) #24 0x0000559f080180fa (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86e0fa) #25 0x000076dcca1d7d90 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29d90) #26 0x000076dcca1d7e40 __libc_start_main (/usr/local/lib/python3.10/dist-packages/tpu_mlir/lib/third_party/libc.so.6+0x29e40) #27 0x0000559f08017505 (/usr/local/lib/python3.10/dist-packages/tpu_mlir/bin/tpuc-opt+0x86d505) Aborted Traceback (most recent call last): File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 556, in <module> lowering_patterns = tool.lowering() File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/tools/model_deploy.py", line 200, in lowering patterns = mlir_lowering( File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 559, in mlir_lowering _os_system(cmd, mute=mute,log_level=log_level) File "/usr/local/lib/python3.10/dist-packages/tpu_mlir/python/utils/mlir_shell.py", line 420, in _os_system raise RuntimeError("[!Error]: {}".format(cmd_str)) RuntimeError: [!Error]: tpuc-opt yolo11n.mlir --processor-assign="chip=cv181x mode=F16 num_device=1 num_core=1 addr_mode=auto high_precision=False" --processor-top-optimize --fuse-preprocess="mode=F16 customization_format=RGB_PACKED align=False" --convert-top-to-tpu=" asymmetric=False doWinograd=False q_group_size=0 matmul_perchannel=False gelu_mode=normal" --canonicalize --weight-fold -o yolo11n_cv181x_f16_tpu.mlir root@25f2c35c9495:/workspace#
Try to downgrade the toolchain version?
pip install tpu_mlir[onnx]==1.14
Hi, it worked with the suggested version of tpu_mlir and removing test_input from deploy command.
Can you update documentation until a new release will fix the error?
Hi, it worked with the suggested version of tpu_mlir and removing test_input from deploy command.
Can you update documentation until a new release will fix the error?
yes, we have updated the wiki. Before fixing the bug, 1.14 version is recommend,
✅ @webfrank
This issue is closed, If you have any questions, you can comment and reply.