AssertionError: Thresholds can't be expressed with type UINT4

[MFJI]

Hello,

I am running the resnet50 model of the finn example in the finn dev branch and I got this error:

Running step: step_hls_codegen [7/13]
Traceback (most recent call last):
File "/home/mengfei/finn/finn/src/finn/builder/build_dataflow.py", line 158, in build_dataflow_cfg
model = transform_step(model, cfg)
File "/home/mengfei/finn/finn/src/finn/builder/build_dataflow_steps.py", line 488, in step_hls_codegen
model = model.transform(PrepareIP(cfg._resolve_fpga_part(), cfg._resolve_hls_clk_period()))
File "/home/mengfei/finn/finn/deps/qonnx/src/qonnx/core/modelwrapper.py", line 140, in transform
(transformed_model, model_was_changed) = transformation.apply(transformed_model)
File "/home/mengfei/finn/finn/src/finn/transformation/fpgadataflow/prepare_ip.py", line 86, in apply
_codegen_single_node(node, model, self.fpgapart, self.clk)
File "/home/mengfei/finn/finn/src/finn/transformation/fpgadataflow/prepare_ip.py", line 53, in _codegen_single_node
inst.code_generation_ipgen(model, fpgapart, clk)
File "/home/mengfei/finn/finn/src/finn/custom_op/fpgadataflow/hlscustomop.py", line 288, in code_generation_ipgen
self.generate_params(model, path)
File "/home/mengfei/finn/finn/src/finn/custom_op/fpgadataflow/thresholding_batch.py", line 445, in generate_params
self.make_weight_file(thresholds, "hls_header", weight_filename)
File "/home/mengfei/finn/finn/src/finn/custom_op/fpgadataflow/thresholding_batch.py", line 350, in make_weight_file
assert np.vectorize(tdt.allowed)(
AssertionError: Thresholds can't be expressed with type UINT4

/home/mengfei/finn/finn/src/finn/custom_op/fpgadataflow/thresholding_batch.py(350)make_weight_file()
-> assert np.vectorize(tdt.allowed)

I didn't change anything in the finn-example. Does anybody know what cause this error and how to fix it? Thank you.

Merry Christmas