mirror of
https://github.com/microsoft/autogen.git
synced 2025-09-14 02:35:40 +00:00
2ebddd67ae
* rm classification head in nlp * rm classification head in nlp * rm classification head in nlp * adding test cases for switch classification head * adding test cases for switch classification head * Update test/nlp/test_autohf_classificationhead.py Co-authored-by: Chi Wang <wang.chi@microsoft.com> * adding test cases for switch classification head * run each test separately * skip classification head test on windows * disabling wandb reporting * fix test nlp custom metric * fix test nlp custom metric * fix test nlp custom metric * fix test nlp custom metric * fix test nlp custom metric * fix test nlp custom metric * fix test nlp custom metric * fix test nlp custom metric * fix test nlp custom metric * fix test nlp custom metric * fix test nlp custom metric * Update website/docs/Examples/AutoML-NLP.md Co-authored-by: Chi Wang <wang.chi@microsoft.com> * Update website/docs/Examples/AutoML-NLP.md Co-authored-by: Chi Wang <wang.chi@microsoft.com> * fix test nlp custom metric Co-authored-by: Chi Wang <wang.chi@microsoft.com>
120 lines
3.3 KiB
Python
120 lines
3.3 KiB
Python
import sys
|
|
import pytest
|
|
import requests
|
|
import os
|
|
import shutil
|
|
from utils import (
|
|
get_toy_data_tokenclassification_idlabel,
|
|
get_toy_data_tokenclassification_tokenlabel,
|
|
get_automl_settings,
|
|
)
|
|
|
|
|
|
@pytest.mark.skipif(
|
|
sys.platform == "darwin" or sys.version < "3.7",
|
|
reason="do not run on mac os or py<3.7",
|
|
)
|
|
def test_tokenclassification_idlabel():
|
|
from flaml import AutoML
|
|
|
|
X_train, y_train, X_val, y_val = get_toy_data_tokenclassification_idlabel()
|
|
automl = AutoML()
|
|
|
|
automl_settings = get_automl_settings()
|
|
automl_settings["task"] = "token-classification"
|
|
automl_settings[
|
|
"metric"
|
|
] = "seqeval:overall_f1" # evaluating based on the overall_f1 of seqeval
|
|
automl_settings["fit_kwargs_by_estimator"]["transformer"]["label_list"] = [
|
|
"O",
|
|
"B-PER",
|
|
"I-PER",
|
|
"B-ORG",
|
|
"I-ORG",
|
|
"B-LOC",
|
|
"I-LOC",
|
|
"B-MISC",
|
|
"I-MISC",
|
|
]
|
|
|
|
try:
|
|
automl.fit(
|
|
X_train=X_train,
|
|
y_train=y_train,
|
|
X_val=X_val,
|
|
y_val=y_val,
|
|
**automl_settings
|
|
)
|
|
except requests.exceptions.HTTPError:
|
|
return
|
|
|
|
# perf test
|
|
import json
|
|
|
|
with open("seqclass.log", "r") as fin:
|
|
for line in fin:
|
|
each_log = json.loads(line.strip("\n"))
|
|
if "validation_loss" in each_log:
|
|
val_loss = each_log["validation_loss"]
|
|
min_inter_result = min(
|
|
each_dict.get("eval_automl_metric", sys.maxsize)
|
|
for each_dict in each_log["logged_metric"]["intermediate_results"]
|
|
)
|
|
|
|
if min_inter_result != sys.maxsize:
|
|
assert val_loss == min_inter_result
|
|
|
|
if os.path.exists("test/data/output/"):
|
|
shutil.rmtree("test/data/output/")
|
|
|
|
|
|
@pytest.mark.skipif(
|
|
sys.platform == "darwin" or sys.version < "3.7",
|
|
reason="do not run on mac os or py<3.7",
|
|
)
|
|
def test_tokenclassification_tokenlabel():
|
|
from flaml import AutoML
|
|
|
|
X_train, y_train, X_val, y_val = get_toy_data_tokenclassification_tokenlabel()
|
|
automl = AutoML()
|
|
|
|
automl_settings = get_automl_settings()
|
|
automl_settings["task"] = "token-classification"
|
|
automl_settings[
|
|
"metric"
|
|
] = "seqeval:overall_f1" # evaluating based on the overall_f1 of seqeval
|
|
|
|
try:
|
|
automl.fit(
|
|
X_train=X_train,
|
|
y_train=y_train,
|
|
X_val=X_val,
|
|
y_val=y_val,
|
|
**automl_settings
|
|
)
|
|
except requests.exceptions.HTTPError:
|
|
return
|
|
|
|
# perf test
|
|
import json
|
|
|
|
with open("seqclass.log", "r") as fin:
|
|
for line in fin:
|
|
each_log = json.loads(line.strip("\n"))
|
|
if "validation_loss" in each_log:
|
|
val_loss = each_log["validation_loss"]
|
|
min_inter_result = min(
|
|
each_dict.get("eval_automl_metric", sys.maxsize)
|
|
for each_dict in each_log["logged_metric"]["intermediate_results"]
|
|
)
|
|
|
|
if min_inter_result != sys.maxsize:
|
|
assert val_loss == min_inter_result
|
|
|
|
if os.path.exists("test/data/output/"):
|
|
shutil.rmtree("test/data/output/")
|
|
|
|
|
|
if __name__ == "__main__":
|
|
test_tokenclassification_idlabel()
|