Problem with resnet50 build for Alveo

[Galapple]

Hi, I am trying to build the design for the Alveo U250 with 3_1 shell.
I have set the enviroment as decribed (I think) but get this error when runing the command:
./run-docker.sh build_custom $FINN_EXAMPLES/build/resnet50

These are the errors:
Traceback (most recent call last):
File "/home/gal/finn_1/finn-examples/build/finn/src/finn/builder/build_dataflow.py", line 168, in build_dataflow_cfg
model = transform_step(model, cfg)
File "/home/gal/finn_1/finn-examples/build/resnet50/custom_steps.py", line 224, in step_resnet50_convert_to_hls
model = model.transform(trn())
File "/home/gal/finn_1/finn-examples/build/finn/deps/qonnx/src/qonnx/core/modelwrapper.py", line 140, in transform
(transformed_model, model_was_changed) = transformation.apply(transformed_model)
File "/home/gal/finn_1/finn-examples/build/finn/src/finn/transformation/fpgadataflow/convert_to_hls_layers.py", line 1243, in apply
model = model.transform(InferShapes())
File "/home/gal/finn_1/finn-examples/build/finn/deps/qonnx/src/qonnx/core/modelwrapper.py", line 140, in transform
(transformed_model, model_was_changed) = transformation.apply(transformed_model)
File "/home/gal/finn_1/finn-examples/build/finn/deps/qonnx/src/qonnx/transformation/infer_shapes.py", line 88, in apply
hidden_ops = _hide_finn_ops(model)
File "/home/gal/finn_1/finn-examples/build/finn/deps/qonnx/src/qonnx/transformation/infer_shapes.py", line 60, in _hide_finn_ops
new_node = _make_shape_compatible_op(node, model)
File "/home/gal/finn_1/finn-examples/build/finn/deps/qonnx/src/qonnx/transformation/infer_shapes.py", line 44, in _make_shape_compatible_op
inst = registry.getCustomOp(node)
File "/home/gal/finn_1/finn-examples/build/finn/deps/qonnx/src/qonnx/custom_op/registry.py", line 45, in getCustomOp
inst = inst_wrapper(node, onnx_opset_version=onnx_opset_version)
TypeError: init() got an unexpected keyword argument 'onnx_opset_version'

/home/gal/finn_1/finn-examples/build/finn/deps/qonnx/src/qonnx/custom_op/registry.py(45)getCustomOp()
-> inst = inst_wrapper(node, onnx_opset_version=onnx_opset_version)
(Pdb)

[mariodruiz]

Can you please described the steps you took?

[Galapple]

Hi Mario, 1. I have created a docker container 2. Clone the FINN and the FINN examples repositories 3. installed conda 4. Changed all env variables as described here https://finn.readthedocs.io/en/latest/getting_started.html#requirements 5. Run this command to build the resnet50 file :./run-docker.sh build_custom $FINN_EXAMPLES/build/resnet50 The build starts and stops at step 7 (step_hls_codegen [7/13]) with these warnings: WARNING: Unable to find /settings64.sh WARNING: Functionality dependent on Vitis will not be available. WARNING: If you need Vitis, ensure VITIS_PATH is set correctly and mounted into the Docker container. WARNING: Unable to find /settings64.sh WARNING: Functionality dependent on Vivado will not be available. WARNING: If you need Vivado, ensure VIVADO_PATH is set correctly and mounted into the Docker container. WARNING: Unable to find /settings64.sh WARNING: Functionality dependent on Vitis HLS will not be available. WARNING: Please note that FINN needs at least version 2020.2 for Vitis HLS support. WARNING: If you need Vitis HLS, ensure HLS_PATH is set correctly and mounted into the Docker container.

…

On Tue, Mar 26, 2024 at 7:30 PM Mario Ruiz ***@***.***> wrote: Can you please described the steps you took? — Reply to this email directly, view it on GitHub <#1013 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AUYYARKFZXXUB5V2URKKIQDY2GPBBAVCNFSM6AAAAABFHHUA5GVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DSMJYGUZTM> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[mariodruiz]

which version of Vitis do you have installed?

[Galapple]

2021.1

…

On Wed, Mar 27, 2024 at 7:13 PM Mario Ruiz ***@***.***> wrote: which version of Vitis do you have installed? — Reply to this email directly, view it on GitHub <#1013 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AUYYARMEN4PKHXWTJWLRDFTY2LVZXAVCNFSM6AAAAABFHHUA5GVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DSMZRGQ2TG> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[mariodruiz]

Hi @Galapple,

By any chance are you running the docker container inside a conda env.
For your information, the development phase should only run on docker, no conda.

For, deployment you will need the docker container.

More advice,

I would recommend that you use finn-examples dev branch and use the get-finn.sh script like described in the README of finn-examples: https://github.com/Xilinx/finn-examples/tree/dev/build
Also, I would suggest you use Vivado 2022.2

[Galapple]

I have tried it with deactivating the conda env and I have got this errors: Running step: step_hls_ipgen [8/13] multiprocessing.pool.RemoteTraceback: """ Traceback (most recent call last): File "/opt/conda/lib/python3.8/multiprocessing/pool.py", line 125, in worker result = (True, func(*args, **kwds)) File "/opt/conda/lib/python3.8/multiprocessing/pool.py", line 48, in mapstar return list(map(*args)) File "/home/gal/finn_1/finn-examples/build/finn/src/finn/transformation/fpgadataflow/hlssynth_ip.py", line 73, in applyNodeLocal inst.ipgen_singlenode_code() File "/home/gal/finn_1/finn-examples/build/finn/src/finn/custom_op/fpgadataflow/hlscustomop.py", line 358, in ipgen_singlenode_code assert os.path.isdir( AssertionError: IPGen failed: /tmp/finn_dev_gal/code_gen_ipgen_ChannelwiseOp_Batch_2_8iis4ywo/project_ChannelwiseOp_Batch_2/sol1/impl/ip not found. Check log under /tmp/finn_dev_gal/code_gen_ipgen_ChannelwiseOp_Batch_2_8iis4ywo """ The above exception was the direct cause of the following exception: Traceback (most recent call last): File "/home/gal/finn_1/finn-examples/build/finn/src/finn/builder/build_dataflow.py", line 168, in build_dataflow_cfg model = transform_step(model, cfg) File "/home/gal/finn_1/finn-examples/build/finn/src/finn/builder/build_dataflow_steps.py", line 493, in step_hls_ipgen model = model.transform(HLSSynthIP()) File "/home/gal/finn_1/finn-examples/build/finn/deps/qonnx/src/qonnx/core/modelwrapper.py", line 140, in transform (transformed_model, model_was_changed) = transformation.apply(transformed_model) File "/home/gal/finn_1/finn-examples/build/finn/deps/qonnx/src/qonnx/transformation/base.py", line 111, in apply new_nodes_and_bool = p.map(self.applyNodeLocal, old_nodes, chunksize=1) File "/opt/conda/lib/python3.8/multiprocessing/pool.py", line 364, in map return self._map_async(func, iterable, mapstar, chunksize).get() File "/opt/conda/lib/python3.8/multiprocessing/pool.py", line 771, in get raise self._value AssertionError: IPGen failed: /tmp/finn_dev_gal/code_gen_ipgen_ChannelwiseOp_Batch_2_8iis4ywo/project_ChannelwiseOp_Batch_2/sol1/impl/ip not found. Check log under /tmp/finn_dev_gal/code_gen_ipgen_ChannelwiseOp_Batch_2_8iis4ywo

/opt/conda/lib/python3.8/multiprocessing/pool.py(771)get()

-> raise self._value Should I upgrade the vitis to 2022.2 or just the vivado?

…

On Thu, Mar 28, 2024 at 4:45 PM Mario Ruiz ***@***.***> wrote: Hi @Galapple <https://github.com/Galapple>, By any chance are you running the docker container inside a conda env. For your information, the development phase should only run on docker, no conda. For, deployment you will need the docker container. More advice, I would recommend that you use finn-examples *dev* branch and use the get-finn.sh script like described in the README of finn-examples: https://github.com/Xilinx/finn-examples/tree/dev/build Also, I would suggest you use Vivado 2022.2 — Reply to this email directly, view it on GitHub <#1013 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AUYYARNHI2AKJPPXKDY67UDY2QNGZAVCNFSM6AAAAABFHHUA5GVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DSNBRGYYDA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[mariodruiz]

Hi @Galapple,

Please start in a new terminal and using the finn-examples dev branch.

Should I upgrade the vitis to 2022.2 or just the vivado?

Vitis