Skip to content

🌟 The Multi-Agent Framework: First AI Software Company, Towards Natural Language Programming

edit-baseline-experimenter

Cyzus Chi9abb8db95496
diff --git a/metagpt/ext/sela/experimenter/aide.py b/metagpt/ext/sela/experimenter/aide.py
index fb71dbdab..50fae94c1 100644
--- a/metagpt/ext/sela/experimenter/aide.py
+++ b/metagpt/ext/sela/experimenter/aide.py
@@ -1,11 +1,15 @@
-import aide
import os
import time
+import aide
+
os.environ["OPENAI_API_KEY"] = "sk-xxx"
os.environ["OPENAI_BASE_URL"] = "your url"
+
start_time = time.time()
+
data_dir = "xxx/data/titanic"
+
goal = f"""
# User requirement
({data_dir}, 'This is a 04_titanic dataset. Your goal is to predict the target column `Survived`.\nPerform data analysis, data preprocessing, feature engineering, and modeling to predict the target. \nReport f1 on the eval data. Do not plot or make any visualizations.\n')
@@ -28,4 +32,4 @@ print(f"Best solution code: {best_solution.code}")
end_time = time.time()
execution_time = end_time - start_time
-print(f"run time : {execution_time} seconds")
\ No newline at end of file
+print(f"run time : {execution_time} seconds")
diff --git a/metagpt/ext/sela/experimenter/autogluon.py b/metagpt/ext/sela/experimenter/autogluon.py
index e7593cfb9..f547ce4ba 100644
--- a/metagpt/ext/sela/experimenter/autogluon.py
+++ b/metagpt/ext/sela/experimenter/autogluon.py
@@ -1,8 +1,10 @@
-from datetime import datetime
-from metagpt.ext.sela.experimenter.custom import CustomExperimenter
import os
+from datetime import datetime
+
import pandas as pd
+from metagpt.ext.sela.experimenter.custom import CustomExperimenter
+
class AGRunner:
def __init__(self, state=None):
@@ -11,6 +13,7 @@ class AGRunner:
def run(self):
from autogluon.tabular import TabularDataset, TabularPredictor
+
train_path = self.datasets["train"]
dev_path = self.datasets["dev"]
dev_wo_target_path = self.datasets["dev_wo_target"]
@@ -32,6 +35,7 @@ class AGRunner:
def run_multimodal(self):
from autogluon.multimodal import MultiModalPredictor
+
target_col = self.state["dataset_config"]["target_col"]
train_path = self.datasets["train"]
dev_path = self.datasets["dev"]
@@ -56,10 +60,7 @@ class AGRunner:
test_preds = predictor.predict(test_data)
# Return predictions for dev and test datasets
- return {
- "dev_preds": dev_preds,
- "test_preds": test_preds
- }
+ return {"dev_preds": dev_preds, "test_preds": test_preds}
def load_split_dataset(self, train_path, dev_path, dev_wo_target_path, test_wo_target_path):
"""
@@ -94,7 +95,8 @@ class AGRunner:
train_data[image_column] = train_data[image_column].apply(lambda x: os.path.join(root_folder, x))
dev_data[image_column] = dev_data[image_column].apply(lambda x: os.path.join(root_folder, x))
dev_wo_target_data[image_column] = dev_wo_target_data[image_column].apply(
- lambda x: os.path.join(root_folder, x))
+ lambda x: os.path.join(root_folder, x)
+ )
test_data[image_column] = test_data[image_column].apply(lambda x: os.path.join(root_folder, x))
return train_data, dev_data, dev_wo_target_data, test_data
@@ -106,7 +108,7 @@ class GluonExperimenter(CustomExperimenter):
def __init__(self, args, **kwargs):
super().__init__(args, **kwargs)
self.framework = AGRunner(self.state)
- self.is_multimodal = args.is_multimodal if hasattr(args, 'is_multimodal') else False
+ self.is_multimodal = args.is_multimodal if hasattr(args, "is_multimodal") else False
async def run_experiment(self):
if not self.is_multimodal:
diff --git a/metagpt/ext/sela/experimenter/autosklearn.py b/metagpt/ext/sela/experimenter/autosklearn.py
index 718e758d4..f6ff267e7 100644
--- a/metagpt/ext/sela/experimenter/autosklearn.py
+++ b/metagpt/ext/sela/experimenter/autosklearn.py
@@ -1,8 +1,10 @@
from datetime import datetime
+from functools import partial
+
import pandas as pd
-from metagpt.ext.sela.experimenter.custom import CustomExperimenter
+
from metagpt.ext.sela.evaluation.evaluation import evaluate_score
-from functools import partial
+from metagpt.ext.sela.experimenter.custom import CustomExperimenter
def custom_scorer(y_true, y_pred, metric_name):
@@ -19,9 +21,7 @@ class ASRunner:
def create_autosklearn_scorer(self, metric_name):
from autosklearn.metrics import make_scorer
- return make_scorer(
- name=metric_name, score_func=partial(custom_scorer, metric_name=metric_name)
- )
+ return make_scorer(name=metric_name, score_func=partial(custom_scorer, metric_name=metric_name))
def run(self):
import autosklearn.classification