pdft.profiling.ProfileReport

pdft.profiling.ProfileReport#

class pdft.profiling.ProfileReport(basis_class: 'str', m: 'int', n: 'int', batch_size: 'int', n_steps: 'int', device: 'str', records: 'list[StepRecord]' = <factory>, setup_s: 'float' = 0.0, total_s: 'float' = 0.0, peak_gb: 'float' = 0.0, trace_dir: 'str | None' = None)[source]#

Bases: object

Parameters:
__init__(basis_class, m, n, batch_size, n_steps, device, records=<factory>, setup_s=0.0, total_s=0.0, peak_gb=0.0, trace_dir=None)#
Parameters:
Return type:

None

Methods

__init__(basis_class, m, n, batch_size, ...)

summary()

to_csv(path)

Attributes

peak_gb

setup_s

total_s

trace_dir

basis_class

m

n

batch_size

n_steps

device

records

basis_class: str#
m: int#
n: int#
batch_size: int#
n_steps: int#
device: str#
records: list[StepRecord]#
setup_s: float = 0.0#
total_s: float = 0.0#
peak_gb: float = 0.0#
trace_dir: str | None = None#
summary()[source]#
Return type:

str

to_csv(path)[source]#
Parameters:

path (str | Path)

Return type:

None