fastmachinelearning / hls4ml

Machine learning on FPGAs using HLS
https://fastmachinelearning.org/hls4ml
Apache License 2.0
1.27k stars 409 forks source link

Error when generate .xo file #616

Closed liuhao-97 closed 2 years ago

liuhao-97 commented 2 years ago

Thank you for your work! When I run the command "hls_model.build(csim=False,synth=True, ecxport=True, bitfile=True)" with the newest branch, I met the error “ERROR: [Common 17-69] Command failed: The '-kernel_xml' or '-kernel_xml_args' must be specified”. I think there are some error when it runs the command "package_xo -force -xo_path xo_files/${myproject}_kernel.xo -kernel_name krnl_rtl -ip_directory hls4ml_IP". I am using U280. My config is ''' cfg = hls4ml.converters.create_config(board='alveo-u280', clock_period= 2, part='xcu280-fsvh2892-2L-e', backend='VivadoAccelerator') cfg['HLSConfig'] = config

cfg['AcceleratorConfig']['Driver'] = 'python' cfg['AcceleratorConfig']['Board'] = 'alveo-u280' cfg['AcceleratorConfig']['Interface'] = 'axi_stream'

cfg['AcceleratorConfig']['Interface'] = 'axi_lite'

cfg['AcceleratorConfig']['Interface'] = 'axi_master'

cfg['AcceleratorConfig']['Precision']['Input'] = 'float' cfg['AcceleratorConfig']['Precision']['Output'] = 'float' cfg['AcceleratorConfig']['Platform'] = 'xilinx_u280_xdma_201920_3'

cfg['IOType']= 'io_stream' cfg['KerasModel'] = model cfg['OutputDir'] = 'test/hls4ml_prj'

print("-----------------------------------") plotting.print_dict(cfg) print("-----------------------------------") hls_model = hls4ml.converters.keras_to_hls(cfg) hls4ml.model.optimizer.get_optimizer('output_rounding_saturation_mode').configure(layers=[]) '''

thesps commented 2 years ago

Which Vivado version are you using? This Alveo bitfile generation currently only works with 2020.1

liuhao-97 commented 2 years ago

Which Vivado version are you using? This Alveo bitfile generation currently only works with 2020.1

Thanks for your reply. I am using 2020.2. I get it.