Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
|
@@ -7,6 +7,9 @@ import pandas as pd
|
|
| 7 |
from evaluate_agent import evaluate_agent, test_questions
|
| 8 |
from LG_agent import BasicAgent
|
| 9 |
# from LG_agent import test_questions, evaluate_agent
|
|
|
|
|
|
|
|
|
|
| 10 |
|
| 11 |
# (Keep Constants as is)
|
| 12 |
# --- Constants ---
|
|
@@ -23,6 +26,29 @@ DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
|
|
| 23 |
# print(f"Agent returning fixed answer: {fixed_answer}")
|
| 24 |
# return fixed_answer
|
| 25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
def run_and_submit_all( profile: gr.OAuthProfile | None):
|
| 27 |
"""
|
| 28 |
Fetches all questions, runs the BasicAgent on them, submits all answers,
|
|
@@ -166,6 +192,8 @@ with gr.Blocks() as demo:
|
|
| 166 |
|
| 167 |
test_eval_button = gr.Button("🔍 Evaluate on Local GAIA Test Set")
|
| 168 |
|
|
|
|
|
|
|
| 169 |
run_button = gr.Button("Run Evaluation & Submit All Answers")
|
| 170 |
|
| 171 |
status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
|
|
@@ -182,6 +210,11 @@ with gr.Blocks() as demo:
|
|
| 182 |
outputs=[status_output, results_table]
|
| 183 |
)
|
| 184 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 185 |
|
| 186 |
if __name__ == "__main__":
|
| 187 |
print("\n" + "-"*30 + " App Starting " + "-"*30)
|
|
|
|
| 7 |
from evaluate_agent import evaluate_agent, test_questions
|
| 8 |
from LG_agent import BasicAgent
|
| 9 |
# from LG_agent import test_questions, evaluate_agent
|
| 10 |
+
import json
|
| 11 |
+
import csv
|
| 12 |
+
from datasets import load_dataset
|
| 13 |
|
| 14 |
# (Keep Constants as is)
|
| 15 |
# --- Constants ---
|
|
|
|
| 26 |
# print(f"Agent returning fixed answer: {fixed_answer}")
|
| 27 |
# return fixed_answer
|
| 28 |
|
| 29 |
+
def download_gaia_to_csv():
|
| 30 |
+
try:
|
| 31 |
+
# Load dataset
|
| 32 |
+
dataset = load_dataset("gaia-benchmark/GAIA", split="dev", trust_remote_code=True)
|
| 33 |
+
|
| 34 |
+
# Define output file
|
| 35 |
+
output_path = "gaia_dev_data.csv"
|
| 36 |
+
|
| 37 |
+
# Extract all keys from the first entry for CSV headers
|
| 38 |
+
keys = dataset[0].keys()
|
| 39 |
+
|
| 40 |
+
# Write to CSV
|
| 41 |
+
with open(output_path, "w", newline="", encoding="utf-8") as f:
|
| 42 |
+
writer = csv.DictWriter(f, fieldnames=keys)
|
| 43 |
+
writer.writeheader()
|
| 44 |
+
for row in dataset:
|
| 45 |
+
writer.writerow(row)
|
| 46 |
+
|
| 47 |
+
return f"✅ GAIA dev dataset saved as CSV at {output_path}"
|
| 48 |
+
except Exception as e:
|
| 49 |
+
return f"❌ Error: {e}"
|
| 50 |
+
|
| 51 |
+
|
| 52 |
def run_and_submit_all( profile: gr.OAuthProfile | None):
|
| 53 |
"""
|
| 54 |
Fetches all questions, runs the BasicAgent on them, submits all answers,
|
|
|
|
| 192 |
|
| 193 |
test_eval_button = gr.Button("🔍 Evaluate on Local GAIA Test Set")
|
| 194 |
|
| 195 |
+
download_db_button = gr.Button("=== Download DB ====")
|
| 196 |
+
|
| 197 |
run_button = gr.Button("Run Evaluation & Submit All Answers")
|
| 198 |
|
| 199 |
status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
|
|
|
|
| 210 |
outputs=[status_output, results_table]
|
| 211 |
)
|
| 212 |
|
| 213 |
+
download_db_button.click(
|
| 214 |
+
fn=download_gaia_to_csv,
|
| 215 |
+
outputs=[status_output]
|
| 216 |
+
)
|
| 217 |
+
|
| 218 |
|
| 219 |
if __name__ == "__main__":
|
| 220 |
print("\n" + "-"*30 + " App Starting " + "-"*30)
|