da4ml.codegen package

Subpackages

• da4ml.codegen.cpp package
  • Submodules
  • da4ml.codegen.cpp.cpp_codegen module
    • cpp_logic_and_bridge_gen()
    • get_io_types()
    • get_typestr_fn()
    • kif_to_hlslib_type()
    • kif_to_vitis_type()
    • output_gen()
    • ssa_gen()
  • da4ml.codegen.cpp.hls_model module
    • HLSModel
      • HLSModel.compile()
      • HLSModel.predict()
      • HLSModel.write()
  • Module contents
    • HLSModel
      • HLSModel.compile()
      • HLSModel.predict()
      • HLSModel.write()
    • cpp_logic_and_bridge_gen()
• da4ml.codegen.verilog package
  • Submodules
  • da4ml.codegen.verilog.comb module
    • comb_logic_gen()
    • output_gen()
    • ssa_gen()
  • da4ml.codegen.verilog.io_wrapper module
    • binder_gen()
    • generate_io_wrapper()
    • hetero_io_map()
  • da4ml.codegen.verilog.pipeline module
    • pipeline_logic_gen()
  • da4ml.codegen.verilog.verilog_model module
    • VerilogModel
      • VerilogModel.compile()
      • VerilogModel.predict()
      • VerilogModel.write()
    • get_io_kifs()
  • Module contents
    • VerilogModel
      • VerilogModel.compile()
      • VerilogModel.predict()
      • VerilogModel.write()
    • binder_gen()
    • comb_logic_gen()
    • generate_io_wrapper()
    • pipeline_logic_gen()

Module contents

class da4ml.codegen.HLSModel(solution: Solution, prj_name: str, path: str | Path, flavor: str = 'vitis', print_latency: bool = True, part_name: str = 'xcvu13p-flga2577-2-e', pragma: Sequence[str] | None = None, clock_period: int = 5, clock_uncertainty: float = 0.1, io_delay_minmax: tuple[float, float] = (0.2, 0.4))

Bases: object

compile(verbose=False, openmp=True, o3: bool = False, clean=True)

Compile the model to a shared object file

Parameters:

• verbose (bool, optional) – Verbose output, by default False

• openmp (bool, optional) – Enable openmp, by default True

• o3 (bool | None, optional) – Turn on -O3 flag, by default False

• clean (bool, optional) – Remove obsolete shared object files, by default True

Raises:

RuntimeError – If compilation fails

predict(data: ndarray[tuple[int, ...], dtype[T]]) → ndarray[tuple[int, ...], dtype[T]]

Run the model on the input data.

Parameters:

data (NDArray[np.floating]) – Input data to the model. The shape is ignored, and the number of samples is determined by the size of the data.

Returns:

Output of the model in shape (n_samples, output_size).

Return type:

NDArray[np.floating]

write()

class da4ml.codegen.VerilogModel(solution: Solution | CascadedSolution, prj_name: str, path: str | Path, latency_cutoff: float = -1, print_latency: bool = True, part_name: str = 'xcvu13p-flga2577-2-e', clock_period: float = 5, clock_uncertainty: float = 0.1, io_delay_minmax: tuple[float, float] = (0.2, 0.4), register_layers: int = 1)

Bases: object

compile(verbose=False, openmp=True, nproc: int | None = None, o3: bool = False, clean=True)

Compile the generated code to a emulator for logic simulation.

Parameters:

• verbose (bool, optional) – Verbose output, by default False

• openmp (bool, optional) – Enable openmp, by default True

• nproc (int | None, optional) – Number of processes to use for compilation, by default None If None, will use the number of CPU cores, but not more than 32.

• o3 (bool | None, optional) – Turn on -O3 flag, by default False

• clean (bool, optional) – Remove obsolete shared object files, by default True

Raises:

RuntimeError – If compilation fails

predict(data: ndarray[tuple[int, ...], dtype[floating]])

Run the model on the input data.

Parameters:

data (NDArray[np.floating]) – Input data to the model. The shape is ignored, and the number of samples is determined by the size of the data.

Returns:

Output of the model in shape (n_samples, output_size).

Return type:

NDArray[np.float64]

write()

da4ml.codegen.binder_gen(csol: CascadedSolution | Solution, module_name: str, II: int = 1, latency_multiplier: int = 1)

da4ml.codegen.comb_logic_gen(sol: Solution, fn_name: str, print_latency: bool = False, timescale: str | None = None)

da4ml.codegen.cpp_logic_and_bridge_gen(sol: Solution, fn_name: str, flavor: str, pragmas: list[str] | None = None, n_indent: int = 4, n_base_indent: int = 0, print_latency: bool = False)

da4ml.codegen.generate_io_wrapper(sol: Solution | CascadedSolution, module_name: str, pipelined: bool = False)

da4ml.codegen.pipeline_logic_gen(csol: CascadedSolution, name: str, print_latency=False, timescale: str | None = '`timescale 1 ns / 1 ps', register_layers: int = 1)