mirror of
https://github.com/pytorch/pytorch.git
synced 2025-10-20 21:14:14 +08:00
b6bdb67f82
Changes by apply order:
1. Replace all `".."` and `os.pardir` usage with `os.path.dirname(...)`.
2. Replace nested `os.path.dirname(os.path.dirname(...))` call with `str(Path(...).parent.parent)`.
3. Reorder `.absolute()` ~/ `.resolve()`~ and `.parent`: always resolve the path first.
`.parent{...}.absolute()` -> `.absolute().parent{...}`
4. Replace chained `.parent x N` with `.parents[${N - 1}]`: the code is easier to read (see 5.)
`.parent.parent.parent.parent` -> `.parents[3]`
5. ~Replace `.parents[${N - 1}]` with `.parents[${N} - 1]`: the code is easier to read and does not introduce any runtime overhead.~
~`.parents[3]` -> `.parents[4 - 1]`~
6. ~Replace `.parents[2 - 1]` with `.parent.parent`: because the code is shorter and easier to read.~
Pull Request resolved: https://github.com/pytorch/pytorch/pull/129374
Approved by: https://github.com/justinchuby, https://github.com/malfet
65 lines
1.9 KiB
Python
65 lines
1.9 KiB
Python
# mypy: ignore-errors
|
|
import sys
|
|
from pathlib import Path
|
|
|
|
import pandas as pd # type: ignore[import-untyped]
|
|
|
|
|
|
sys.path.append(str(Path(__file__).absolute().parents[1]))
|
|
|
|
from train_decision import AHTrainDecisionTree
|
|
|
|
from torch._inductor.autoheuristic.autoheuristic_utils import mm_operations
|
|
|
|
|
|
class AHTrainDecisionTreeMM(AHTrainDecisionTree):
|
|
def __init__(self):
|
|
super().__init__()
|
|
|
|
def add_new_features(self, results):
|
|
ops = mm_operations()
|
|
added_categorical_features = []
|
|
for op in ops:
|
|
results[op.name] = results.apply(op.func, axis=1)
|
|
if op.is_categorical:
|
|
added_categorical_features.append(op.name)
|
|
return (results, added_categorical_features)
|
|
|
|
def get_default_config(self, row):
|
|
return "extern_mm"
|
|
|
|
def get_allowed_wrong_prediction_pct(self):
|
|
return 1.0
|
|
|
|
def get_test_and_val_size(self):
|
|
return (0.01, 0.19)
|
|
|
|
def get_grid_search_values(self):
|
|
return {"max_depth": [5], "min_samples_leaf": [0.01], "criterion": ["entropy"]}
|
|
|
|
def add_training_data(self, df_train, datasets):
|
|
# add each dataset to the training data 3 times
|
|
# we really want to make sure that the heuristic performs well on these datasets
|
|
df_timm_train = datasets["train_timm"]
|
|
df_timm_train = df_timm_train.loc[df_timm_train.index.repeat(3)].reset_index(
|
|
drop=True
|
|
)
|
|
df_hf_train = datasets["train_hf"]
|
|
df_hf_train = df_hf_train.loc[df_hf_train.index.repeat(3)].reset_index(
|
|
drop=True
|
|
)
|
|
df_train = datasets["train"]
|
|
df_train = pd.concat(
|
|
[df_train, df_timm_train, df_hf_train],
|
|
ignore_index=True,
|
|
)
|
|
return df_train
|
|
|
|
def ranking_always_included_choices(self):
|
|
return ["extern_mm"]
|
|
|
|
|
|
if __name__ == "__main__":
|
|
train = AHTrainDecisionTreeMM()
|
|
train.generate_heuristic()
|