Yuxuan Zhang
commited on
Upload folder using huggingface_hub
Browse files- .gitattributes +6 -0
- .github/workflows/update_space.yml +28 -0
- .gradio/certificate.pem +31 -0
- 2048_performance.png +3 -0
- README.md +3 -9
- __pycache__/data_analysis.cpython-310.pyc +0 -0
- __pycache__/data_visualization.cpython-310.pyc +0 -0
- __pycache__/leaderboard_utils.cpython-310.pyc +0 -0
- assets/2048/2048.gif +3 -0
- assets/candy/candy.gif +3 -0
- assets/sokoban/sokoban.gif +3 -0
- assets/super_mario_bros/super_mario.gif +3 -0
- assets/tetris/tetris.gif +3 -0
- data_analysis.ipynb +1 -0
- data_visualization.py +384 -0
- gradio_app.py +505 -0
- gradio_app_v2.py +540 -0
- leaderboard_utils.py +202 -0
- rank_data_03_25_2025.json +324 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,9 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
2048_performance.png filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
assets/2048/2048.gif filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
assets/candy/candy.gif filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
assets/sokoban/sokoban.gif filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
assets/super_mario_bros/super_mario.gif filter=lfs diff=lfs merge=lfs -text
|
| 41 |
+
assets/tetris/tetris.gif filter=lfs diff=lfs merge=lfs -text
|
.github/workflows/update_space.yml
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
name: Run Python script
|
| 2 |
+
|
| 3 |
+
on:
|
| 4 |
+
push:
|
| 5 |
+
branches:
|
| 6 |
+
- gradio_app_v2_dev
|
| 7 |
+
|
| 8 |
+
jobs:
|
| 9 |
+
build:
|
| 10 |
+
runs-on: ubuntu-latest
|
| 11 |
+
|
| 12 |
+
steps:
|
| 13 |
+
- name: Checkout
|
| 14 |
+
uses: actions/checkout@v2
|
| 15 |
+
|
| 16 |
+
- name: Set up Python
|
| 17 |
+
uses: actions/setup-python@v2
|
| 18 |
+
with:
|
| 19 |
+
python-version: '3.9'
|
| 20 |
+
|
| 21 |
+
- name: Install Gradio
|
| 22 |
+
run: python -m pip install gradio
|
| 23 |
+
|
| 24 |
+
- name: Log in to Hugging Face
|
| 25 |
+
run: python -c 'import huggingface_hub; huggingface_hub.login(token="${{ secrets.hf_token }}")'
|
| 26 |
+
|
| 27 |
+
- name: Deploy to Spaces
|
| 28 |
+
run: gradio deploy
|
.gradio/certificate.pem
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
-----BEGIN CERTIFICATE-----
|
| 2 |
+
MIIFazCCA1OgAwIBAgIRAIIQz7DSQONZRGPgu2OCiwAwDQYJKoZIhvcNAQELBQAw
|
| 3 |
+
TzELMAkGA1UEBhMCVVMxKTAnBgNVBAoTIEludGVybmV0IFNlY3VyaXR5IFJlc2Vh
|
| 4 |
+
cmNoIEdyb3VwMRUwEwYDVQQDEwxJU1JHIFJvb3QgWDEwHhcNMTUwNjA0MTEwNDM4
|
| 5 |
+
WhcNMzUwNjA0MTEwNDM4WjBPMQswCQYDVQQGEwJVUzEpMCcGA1UEChMgSW50ZXJu
|
| 6 |
+
ZXQgU2VjdXJpdHkgUmVzZWFyY2ggR3JvdXAxFTATBgNVBAMTDElTUkcgUm9vdCBY
|
| 7 |
+
MTCCAiIwDQYJKoZIhvcNAQEBBQADggIPADCCAgoCggIBAK3oJHP0FDfzm54rVygc
|
| 8 |
+
h77ct984kIxuPOZXoHj3dcKi/vVqbvYATyjb3miGbESTtrFj/RQSa78f0uoxmyF+
|
| 9 |
+
0TM8ukj13Xnfs7j/EvEhmkvBioZxaUpmZmyPfjxwv60pIgbz5MDmgK7iS4+3mX6U
|
| 10 |
+
A5/TR5d8mUgjU+g4rk8Kb4Mu0UlXjIB0ttov0DiNewNwIRt18jA8+o+u3dpjq+sW
|
| 11 |
+
T8KOEUt+zwvo/7V3LvSye0rgTBIlDHCNAymg4VMk7BPZ7hm/ELNKjD+Jo2FR3qyH
|
| 12 |
+
B5T0Y3HsLuJvW5iB4YlcNHlsdu87kGJ55tukmi8mxdAQ4Q7e2RCOFvu396j3x+UC
|
| 13 |
+
B5iPNgiV5+I3lg02dZ77DnKxHZu8A/lJBdiB3QW0KtZB6awBdpUKD9jf1b0SHzUv
|
| 14 |
+
KBds0pjBqAlkd25HN7rOrFleaJ1/ctaJxQZBKT5ZPt0m9STJEadao0xAH0ahmbWn
|
| 15 |
+
OlFuhjuefXKnEgV4We0+UXgVCwOPjdAvBbI+e0ocS3MFEvzG6uBQE3xDk3SzynTn
|
| 16 |
+
jh8BCNAw1FtxNrQHusEwMFxIt4I7mKZ9YIqioymCzLq9gwQbooMDQaHWBfEbwrbw
|
| 17 |
+
qHyGO0aoSCqI3Haadr8faqU9GY/rOPNk3sgrDQoo//fb4hVC1CLQJ13hef4Y53CI
|
| 18 |
+
rU7m2Ys6xt0nUW7/vGT1M0NPAgMBAAGjQjBAMA4GA1UdDwEB/wQEAwIBBjAPBgNV
|
| 19 |
+
HRMBAf8EBTADAQH/MB0GA1UdDgQWBBR5tFnme7bl5AFzgAiIyBpY9umbbjANBgkq
|
| 20 |
+
hkiG9w0BAQsFAAOCAgEAVR9YqbyyqFDQDLHYGmkgJykIrGF1XIpu+ILlaS/V9lZL
|
| 21 |
+
ubhzEFnTIZd+50xx+7LSYK05qAvqFyFWhfFQDlnrzuBZ6brJFe+GnY+EgPbk6ZGQ
|
| 22 |
+
3BebYhtF8GaV0nxvwuo77x/Py9auJ/GpsMiu/X1+mvoiBOv/2X/qkSsisRcOj/KK
|
| 23 |
+
NFtY2PwByVS5uCbMiogziUwthDyC3+6WVwW6LLv3xLfHTjuCvjHIInNzktHCgKQ5
|
| 24 |
+
ORAzI4JMPJ+GslWYHb4phowim57iaztXOoJwTdwJx4nLCgdNbOhdjsnvzqvHu7Ur
|
| 25 |
+
TkXWStAmzOVyyghqpZXjFaH3pO3JLF+l+/+sKAIuvtd7u+Nxe5AW0wdeRlN8NwdC
|
| 26 |
+
jNPElpzVmbUq4JUagEiuTDkHzsxHpFKVK7q4+63SM1N95R1NbdWhscdCb+ZAJzVc
|
| 27 |
+
oyi3B43njTOQ5yOf+1CceWxG1bQVs5ZufpsMljq4Ui0/1lvh+wjChP4kqKOJ2qxq
|
| 28 |
+
4RgqsahDYVvTH9w7jXbyLeiNdd8XM2w9U/t7y0Ff/9yi0GE44Za4rF2LN9d11TPA
|
| 29 |
+
mRGunUHBcnWEvgJBQl9nJEiU0Zsnvgc/ubhPgXRR4Xq37Z0j4r7g1SgEEzwxA57d
|
| 30 |
+
emyPxgcYxn/eR44/KJ4EBs+lVDR3veyJm+kXQ99b21/+jh5Xos1AnX5iItreGCc=
|
| 31 |
+
-----END CERTIFICATE-----
|
2048_performance.png
ADDED
|
Git LFS Details
|
README.md
CHANGED
|
@@ -1,12 +1,6 @@
|
|
| 1 |
---
|
| 2 |
-
title:
|
| 3 |
-
|
| 4 |
-
colorFrom: yellow
|
| 5 |
-
colorTo: gray
|
| 6 |
sdk: gradio
|
| 7 |
-
sdk_version: 5.23.
|
| 8 |
-
app_file: app.py
|
| 9 |
-
pinned: false
|
| 10 |
---
|
| 11 |
-
|
| 12 |
-
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
|
|
|
| 1 |
---
|
| 2 |
+
title: lmgame
|
| 3 |
+
app_file: gradio_app_v2.py
|
|
|
|
|
|
|
| 4 |
sdk: gradio
|
| 5 |
+
sdk_version: 5.23.1
|
|
|
|
|
|
|
| 6 |
---
|
|
|
|
|
|
__pycache__/data_analysis.cpython-310.pyc
ADDED
|
Binary file (5.34 kB). View file
|
|
|
__pycache__/data_visualization.cpython-310.pyc
ADDED
|
Binary file (10.5 kB). View file
|
|
|
__pycache__/leaderboard_utils.cpython-310.pyc
ADDED
|
Binary file (4.46 kB). View file
|
|
|
assets/2048/2048.gif
ADDED
|
Git LFS Details
|
assets/candy/candy.gif
ADDED
|
Git LFS Details
|
assets/sokoban/sokoban.gif
ADDED
|
Git LFS Details
|
assets/super_mario_bros/super_mario.gif
ADDED
|
Git LFS Details
|
assets/tetris/tetris.gif
ADDED
|
Git LFS Details
|
data_analysis.ipynb
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
|
data_visualization.py
ADDED
|
@@ -0,0 +1,384 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import matplotlib
|
| 2 |
+
matplotlib.use('Agg') # Use Agg backend for thread safety
|
| 3 |
+
import matplotlib.pyplot as plt
|
| 4 |
+
import numpy as np
|
| 5 |
+
import pandas as pd
|
| 6 |
+
from leaderboard_utils import (
|
| 7 |
+
get_organization,
|
| 8 |
+
get_mario_leaderboard,
|
| 9 |
+
get_sokoban_leaderboard,
|
| 10 |
+
get_2048_leaderboard,
|
| 11 |
+
get_candy_leaderboard,
|
| 12 |
+
get_tetris_leaderboard,
|
| 13 |
+
get_tetris_planning_leaderboard,
|
| 14 |
+
get_combined_leaderboard,
|
| 15 |
+
GAME_ORDER
|
| 16 |
+
)
|
| 17 |
+
|
| 18 |
+
# Define game score columns mapping
|
| 19 |
+
GAME_SCORE_COLUMNS = {
|
| 20 |
+
"Super Mario Bros": "Score",
|
| 21 |
+
"Sokoban": "Levels Cracked",
|
| 22 |
+
"2048": "Score",
|
| 23 |
+
"Candy Crash": "Average Score",
|
| 24 |
+
"Tetris (complete)": "Score",
|
| 25 |
+
"Tetris (planning only)": "Score"
|
| 26 |
+
}
|
| 27 |
+
|
| 28 |
+
def simplify_model_name(model_name):
|
| 29 |
+
"""
|
| 30 |
+
Simplify model name by either taking first 11 chars or string before third '-'
|
| 31 |
+
"""
|
| 32 |
+
hyphen_parts = model_name.split('-')
|
| 33 |
+
return '-'.join(hyphen_parts[:3]) if len(hyphen_parts) >= 3 else model_name[:11]
|
| 34 |
+
|
| 35 |
+
def create_horizontal_bar_chart(df, game_name):
|
| 36 |
+
"""
|
| 37 |
+
Create horizontal bar chart for detailed game view
|
| 38 |
+
|
| 39 |
+
Args:
|
| 40 |
+
df (pd.DataFrame): DataFrame containing game data
|
| 41 |
+
game_name (str): Name of the game to display
|
| 42 |
+
|
| 43 |
+
Returns:
|
| 44 |
+
matplotlib.figure.Figure: The generated bar chart figure
|
| 45 |
+
"""
|
| 46 |
+
# Close any existing figures to prevent memory leaks
|
| 47 |
+
plt.close('all')
|
| 48 |
+
|
| 49 |
+
# Set style
|
| 50 |
+
plt.style.use('default')
|
| 51 |
+
# Increase figure width to accommodate long model names
|
| 52 |
+
fig, ax = plt.subplots(figsize=(10, 6))
|
| 53 |
+
|
| 54 |
+
# Sort by score
|
| 55 |
+
if game_name == "Super Mario Bros":
|
| 56 |
+
score_col = "Score"
|
| 57 |
+
df_sorted = df.sort_values(by=score_col, ascending=True)
|
| 58 |
+
elif game_name == "Sokoban":
|
| 59 |
+
# Process Sokoban scores by splitting and getting max level
|
| 60 |
+
def get_max_level(levels_str):
|
| 61 |
+
try:
|
| 62 |
+
# Split by semicolon, strip whitespace, filter empty strings, convert to integers
|
| 63 |
+
levels = [int(x.strip()) for x in levels_str.split(";") if x.strip()]
|
| 64 |
+
return max(levels) if levels else 0
|
| 65 |
+
except:
|
| 66 |
+
return 0
|
| 67 |
+
|
| 68 |
+
# Create a temporary column with max levels
|
| 69 |
+
df['Max Level'] = df['Levels Cracked'].apply(get_max_level)
|
| 70 |
+
df_sorted = df.sort_values(by='Max Level', ascending=True)
|
| 71 |
+
score_col = 'Max Level'
|
| 72 |
+
elif game_name == "2048":
|
| 73 |
+
score_col = "Score"
|
| 74 |
+
df_sorted = df.sort_values(by=score_col, ascending=True)
|
| 75 |
+
elif game_name == "Candy Crash":
|
| 76 |
+
score_col = "Average Score"
|
| 77 |
+
df_sorted = df.sort_values(by=score_col, ascending=True)
|
| 78 |
+
elif game_name in ["Tetris (complete)", "Tetris (planning only)"]:
|
| 79 |
+
score_col = "Score"
|
| 80 |
+
df_sorted = df.sort_values(by=score_col, ascending=True)
|
| 81 |
+
else:
|
| 82 |
+
return None
|
| 83 |
+
|
| 84 |
+
# Create color gradient
|
| 85 |
+
colors = plt.cm.viridis(np.linspace(0.2, 0.8, len(df_sorted)))
|
| 86 |
+
|
| 87 |
+
# Create horizontal bars
|
| 88 |
+
bars = ax.barh(range(len(df_sorted)), df_sorted[score_col], color=colors)
|
| 89 |
+
|
| 90 |
+
# Add more space for labels on the left
|
| 91 |
+
plt.subplots_adjust(left=0.3)
|
| 92 |
+
|
| 93 |
+
# Customize the chart
|
| 94 |
+
ax.set_yticks(range(len(df_sorted)))
|
| 95 |
+
|
| 96 |
+
# Format player names: keep organization info and truncate the rest if too long
|
| 97 |
+
def format_player_name(player, org):
|
| 98 |
+
max_length = 40 # Maximum length for player name
|
| 99 |
+
if len(player) > max_length:
|
| 100 |
+
# Keep the first part and last part of the name
|
| 101 |
+
parts = player.split('-')
|
| 102 |
+
if len(parts) > 3:
|
| 103 |
+
formatted = f"{parts[0]}-{parts[1]}-...{parts[-1]}"
|
| 104 |
+
else:
|
| 105 |
+
formatted = player[:max_length-3] + "..."
|
| 106 |
+
else:
|
| 107 |
+
formatted = player
|
| 108 |
+
return f"{formatted} [{org}]"
|
| 109 |
+
|
| 110 |
+
player_labels = [format_player_name(row['Player'], row['Organization'])
|
| 111 |
+
for _, row in df_sorted.iterrows()]
|
| 112 |
+
ax.set_yticklabels(player_labels, fontsize=9)
|
| 113 |
+
|
| 114 |
+
# Add value labels on the bars
|
| 115 |
+
for i, bar in enumerate(bars):
|
| 116 |
+
width = bar.get_width()
|
| 117 |
+
if game_name == "Candy Crash":
|
| 118 |
+
score_text = f'{width:.1f}'
|
| 119 |
+
else:
|
| 120 |
+
score_text = f'{width:.0f}'
|
| 121 |
+
|
| 122 |
+
ax.text(width, bar.get_y() + bar.get_height()/2,
|
| 123 |
+
score_text,
|
| 124 |
+
ha='left', va='center',
|
| 125 |
+
fontsize=10,
|
| 126 |
+
fontweight='bold',
|
| 127 |
+
color='white',
|
| 128 |
+
bbox=dict(facecolor=(0, 0, 0, 0.3),
|
| 129 |
+
edgecolor='none',
|
| 130 |
+
alpha=0.5,
|
| 131 |
+
pad=2))
|
| 132 |
+
|
| 133 |
+
# Set title and labels
|
| 134 |
+
ax.set_title(f"{game_name} Performance",
|
| 135 |
+
pad=20,
|
| 136 |
+
fontsize=14,
|
| 137 |
+
fontweight='bold',
|
| 138 |
+
color='#2c3e50')
|
| 139 |
+
|
| 140 |
+
if game_name == "Sokoban":
|
| 141 |
+
ax.set_xlabel("Maximum Level Reached",
|
| 142 |
+
fontsize=12,
|
| 143 |
+
fontweight='bold',
|
| 144 |
+
color='#2c3e50',
|
| 145 |
+
labelpad=10)
|
| 146 |
+
else:
|
| 147 |
+
ax.set_xlabel(score_col,
|
| 148 |
+
fontsize=12,
|
| 149 |
+
fontweight='bold',
|
| 150 |
+
color='#2c3e50',
|
| 151 |
+
labelpad=10)
|
| 152 |
+
|
| 153 |
+
# Add grid lines
|
| 154 |
+
ax.grid(True, axis='x', linestyle='--', alpha=0.3)
|
| 155 |
+
|
| 156 |
+
# Remove top and right spines
|
| 157 |
+
ax.spines['top'].set_visible(False)
|
| 158 |
+
ax.spines['right'].set_visible(False)
|
| 159 |
+
|
| 160 |
+
# Adjust layout
|
| 161 |
+
plt.tight_layout()
|
| 162 |
+
|
| 163 |
+
return fig
|
| 164 |
+
|
| 165 |
+
def create_radar_charts(df):
|
| 166 |
+
"""
|
| 167 |
+
Create two radar charts with improved normalization using z-scores
|
| 168 |
+
"""
|
| 169 |
+
# Close any existing figures to prevent memory leaks
|
| 170 |
+
plt.close('all')
|
| 171 |
+
|
| 172 |
+
# Define reasoning models
|
| 173 |
+
reasoning_models = [
|
| 174 |
+
'claude-3-7-sonnet-20250219(thinking)',
|
| 175 |
+
'o1-2024-12-17',
|
| 176 |
+
'gemini-2.0-flash-thinking-exp-1219',
|
| 177 |
+
'o3-mini-2025-01-31(medium)',
|
| 178 |
+
'gemini-2.5-pro-exp-03-25',
|
| 179 |
+
'o1-mini-2024-09-12',
|
| 180 |
+
'deepseek-r1'
|
| 181 |
+
]
|
| 182 |
+
|
| 183 |
+
# Split dataframe into reasoning and non-reasoning models
|
| 184 |
+
df_reasoning = df[df['Player'].isin(reasoning_models)]
|
| 185 |
+
df_others = df[~df['Player'].isin(reasoning_models)]
|
| 186 |
+
|
| 187 |
+
# Get game columns
|
| 188 |
+
game_columns = [col for col in df.columns if col.endswith(' Score')]
|
| 189 |
+
categories = [col.replace(' Score', '') for col in game_columns]
|
| 190 |
+
|
| 191 |
+
# Create figure with two subplots - adjusted size for new layout
|
| 192 |
+
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(10, 6), subplot_kw=dict(projection='polar'))
|
| 193 |
+
fig.patch.set_facecolor('white') # Set figure background to white
|
| 194 |
+
|
| 195 |
+
def normalize_values(values, mean, std):
|
| 196 |
+
"""
|
| 197 |
+
Normalize values using z-score and scale to 0-100 range
|
| 198 |
+
"""
|
| 199 |
+
if std == 0:
|
| 200 |
+
return [50 if v > 0 else 0 for v in values] # Handle zero std case
|
| 201 |
+
z_scores = [(v - mean) / std for v in values]
|
| 202 |
+
# Scale z-scores to 0-100 range, with mean at 50
|
| 203 |
+
scaled_values = [max(0, min(100, (z * 30) + 30)) for z in z_scores]
|
| 204 |
+
return scaled_values
|
| 205 |
+
|
| 206 |
+
def get_game_stats(df, game_col):
|
| 207 |
+
"""
|
| 208 |
+
Get mean and std for a game column, handling missing values
|
| 209 |
+
"""
|
| 210 |
+
values = []
|
| 211 |
+
for val in df[game_col]:
|
| 212 |
+
if isinstance(val, str) and val == '_':
|
| 213 |
+
values.append(0)
|
| 214 |
+
else:
|
| 215 |
+
try:
|
| 216 |
+
values.append(float(val))
|
| 217 |
+
except:
|
| 218 |
+
values.append(0)
|
| 219 |
+
return np.mean(values), np.std(values)
|
| 220 |
+
|
| 221 |
+
def setup_radar_plot(ax, data, title):
|
| 222 |
+
ax.set_facecolor('white') # Set subplot background to white
|
| 223 |
+
|
| 224 |
+
num_vars = len(categories)
|
| 225 |
+
angles = np.linspace(0, 2*np.pi, num_vars, endpoint=False)
|
| 226 |
+
angles = np.concatenate((angles, [angles[0]]))
|
| 227 |
+
|
| 228 |
+
# Plot grid lines with darker color
|
| 229 |
+
grid_values = [10, 30, 50, 70, 90]
|
| 230 |
+
ax.set_rgrids(grid_values,
|
| 231 |
+
labels=grid_values,
|
| 232 |
+
angle=45,
|
| 233 |
+
fontsize=6,
|
| 234 |
+
alpha=0.7, # Increased alpha for better visibility
|
| 235 |
+
color='#404040') # Darker color for grid labels
|
| 236 |
+
|
| 237 |
+
# Make grid lines darker but still subtle
|
| 238 |
+
ax.grid(True, color='#404040', alpha=0.3) # Darker grid lines
|
| 239 |
+
|
| 240 |
+
# Define darker, more vibrant colors for the radar plots
|
| 241 |
+
colors = ['#1f77b4', '#d62728', '#2ca02c', '#ff7f0e', '#9467bd', '#8c564b']
|
| 242 |
+
|
| 243 |
+
# Calculate game statistics once
|
| 244 |
+
game_stats = {col: get_game_stats(df, col) for col in game_columns}
|
| 245 |
+
|
| 246 |
+
# Plot data with darker lines and higher opacity for fills
|
| 247 |
+
for idx, (_, row) in enumerate(data.iterrows()):
|
| 248 |
+
values = []
|
| 249 |
+
for col in game_columns:
|
| 250 |
+
val = row[col]
|
| 251 |
+
if isinstance(val, str) and val == '_':
|
| 252 |
+
values.append(0)
|
| 253 |
+
else:
|
| 254 |
+
try:
|
| 255 |
+
values.append(float(val))
|
| 256 |
+
except:
|
| 257 |
+
values.append(0)
|
| 258 |
+
|
| 259 |
+
# Normalize values using game statistics
|
| 260 |
+
normalized_values = []
|
| 261 |
+
for i, v in enumerate(values):
|
| 262 |
+
mean, std = game_stats[game_columns[i]]
|
| 263 |
+
normalized_value = normalize_values([v], mean, std)[0]
|
| 264 |
+
normalized_values.append(normalized_value)
|
| 265 |
+
|
| 266 |
+
# Complete the circular plot
|
| 267 |
+
normalized_values = np.concatenate((normalized_values, [normalized_values[0]]))
|
| 268 |
+
|
| 269 |
+
model_name = simplify_model_name(row['Player'])
|
| 270 |
+
ax.plot(angles, normalized_values, 'o-', linewidth=2.0, # Increased line width
|
| 271 |
+
label=model_name,
|
| 272 |
+
color=colors[idx % len(colors)],
|
| 273 |
+
markersize=4) # Increased marker size
|
| 274 |
+
ax.fill(angles, normalized_values,
|
| 275 |
+
alpha=0.3, # Increased fill opacity
|
| 276 |
+
color=colors[idx % len(colors)])
|
| 277 |
+
|
| 278 |
+
# Format categories
|
| 279 |
+
formatted_categories = []
|
| 280 |
+
for game in categories:
|
| 281 |
+
if game == "Tetris (planning only)":
|
| 282 |
+
game = "Tetris\n(planning)"
|
| 283 |
+
elif game == "Tetris (complete)":
|
| 284 |
+
game = "Tetris\n(complete)"
|
| 285 |
+
elif game == "Super Mario Bros":
|
| 286 |
+
game = "Super\nMario"
|
| 287 |
+
elif game == "Candy Crash":
|
| 288 |
+
game = "Candy\nCrash"
|
| 289 |
+
formatted_categories.append(game)
|
| 290 |
+
|
| 291 |
+
ax.set_xticks(angles[:-1])
|
| 292 |
+
ax.set_xticklabels(formatted_categories,
|
| 293 |
+
fontsize=8, # Slightly larger font
|
| 294 |
+
color='#202020', # Darker text
|
| 295 |
+
fontweight='bold') # Bold text
|
| 296 |
+
ax.tick_params(pad=10, colors='#202020') # Darker tick colors
|
| 297 |
+
|
| 298 |
+
ax.set_title(title,
|
| 299 |
+
pad=20,
|
| 300 |
+
fontsize=11, # Slightly larger title
|
| 301 |
+
color='#202020', # Darker title
|
| 302 |
+
fontweight='bold') # Bold title
|
| 303 |
+
|
| 304 |
+
legend = ax.legend(loc='upper right',
|
| 305 |
+
bbox_to_anchor=(1.3, 1.1),
|
| 306 |
+
fontsize=7, # Slightly larger legend
|
| 307 |
+
framealpha=0.9, # More opaque legend
|
| 308 |
+
edgecolor='#404040', # Darker edge
|
| 309 |
+
ncol=1)
|
| 310 |
+
|
| 311 |
+
ax.set_ylim(0, 105)
|
| 312 |
+
ax.spines['polar'].set_color('#404040') # Darker spine
|
| 313 |
+
ax.spines['polar'].set_alpha(0.5) # More visible spine
|
| 314 |
+
|
| 315 |
+
# Setup both plots
|
| 316 |
+
setup_radar_plot(ax1, df_reasoning, "Reasoning Models")
|
| 317 |
+
setup_radar_plot(ax2, df_others, "Non-Reasoning Models")
|
| 318 |
+
|
| 319 |
+
plt.subplots_adjust(right=0.85, wspace=0.3)
|
| 320 |
+
|
| 321 |
+
return fig
|
| 322 |
+
|
| 323 |
+
def get_combined_leaderboard_with_radar(rank_data, selected_games):
|
| 324 |
+
"""
|
| 325 |
+
Get combined leaderboard and create radar charts
|
| 326 |
+
"""
|
| 327 |
+
df = get_combined_leaderboard(rank_data, selected_games)
|
| 328 |
+
radar_fig = create_radar_charts(df)
|
| 329 |
+
return df, radar_fig
|
| 330 |
+
|
| 331 |
+
def create_organization_radar_chart(rank_data):
|
| 332 |
+
"""
|
| 333 |
+
Create radar chart comparing organizations
|
| 334 |
+
"""
|
| 335 |
+
# Get combined leaderboard with all games
|
| 336 |
+
df = get_combined_leaderboard(rank_data, {game: True for game in GAME_ORDER})
|
| 337 |
+
|
| 338 |
+
# Group by organization and calculate average scores
|
| 339 |
+
org_performance = {}
|
| 340 |
+
for org in df["Organization"].unique():
|
| 341 |
+
org_df = df[df["Organization"] == org]
|
| 342 |
+
scores = {}
|
| 343 |
+
for game in GAME_ORDER:
|
| 344 |
+
game_scores = org_df[f"{game} Score"].apply(lambda x: float(x) if x != "_" else 0)
|
| 345 |
+
scores[game] = game_scores.mean()
|
| 346 |
+
org_performance[org] = scores
|
| 347 |
+
|
| 348 |
+
# Create radar chart
|
| 349 |
+
return create_radar_charts(pd.DataFrame([org_performance]))
|
| 350 |
+
|
| 351 |
+
def create_top_players_radar_chart(rank_data, n=5):
|
| 352 |
+
"""
|
| 353 |
+
Create radar chart for top N players
|
| 354 |
+
"""
|
| 355 |
+
# Get combined leaderboard with all games
|
| 356 |
+
df = get_combined_leaderboard(rank_data, {game: True for game in GAME_ORDER})
|
| 357 |
+
|
| 358 |
+
# Get top N players
|
| 359 |
+
top_players = df["Player"].head(n).tolist()
|
| 360 |
+
|
| 361 |
+
# Create radar chart for top players
|
| 362 |
+
return create_radar_charts(df[df["Player"].isin(top_players)])
|
| 363 |
+
|
| 364 |
+
def create_player_radar_chart(rank_data, player_name):
|
| 365 |
+
"""
|
| 366 |
+
Create radar chart for a specific player
|
| 367 |
+
"""
|
| 368 |
+
# Get combined leaderboard with all games
|
| 369 |
+
df = get_combined_leaderboard(rank_data, {game: True for game in GAME_ORDER})
|
| 370 |
+
|
| 371 |
+
# Get player's data
|
| 372 |
+
player_df = df[df["Player"] == player_name]
|
| 373 |
+
|
| 374 |
+
if player_df.empty:
|
| 375 |
+
return None
|
| 376 |
+
|
| 377 |
+
# Create radar chart for the player
|
| 378 |
+
return create_radar_charts(player_df)
|
| 379 |
+
|
| 380 |
+
def save_visualization(fig, filename):
|
| 381 |
+
"""
|
| 382 |
+
Save visualization to file
|
| 383 |
+
"""
|
| 384 |
+
fig.savefig(filename, bbox_inches='tight', dpi=300)
|
gradio_app.py
ADDED
|
@@ -0,0 +1,505 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import os
|
| 3 |
+
import pandas as pd
|
| 4 |
+
from PIL import Image, ImageSequence
|
| 5 |
+
import io
|
| 6 |
+
|
| 7 |
+
#######################################################
|
| 8 |
+
# Red Baron Game HTML
|
| 9 |
+
#######################################################
|
| 10 |
+
test_html = """
|
| 11 |
+
<!DOCTYPE html>
|
| 12 |
+
<html lang="en">
|
| 13 |
+
<head>
|
| 14 |
+
<meta charset="UTF-8">
|
| 15 |
+
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
| 16 |
+
<title>Simple Test Page</title>
|
| 17 |
+
<style>
|
| 18 |
+
body {
|
| 19 |
+
font-family: Arial, sans-serif;
|
| 20 |
+
margin: 0;
|
| 21 |
+
padding: 20px;
|
| 22 |
+
background-color: #f0f8ff;
|
| 23 |
+
color: #333;
|
| 24 |
+
}
|
| 25 |
+
.container {
|
| 26 |
+
max-width: 800px;
|
| 27 |
+
margin: 0 auto;
|
| 28 |
+
padding: 20px;
|
| 29 |
+
background-color: white;
|
| 30 |
+
border-radius: 8px;
|
| 31 |
+
box-shadow: 0 2px 4px rgba(0,0,0,0.1);
|
| 32 |
+
}
|
| 33 |
+
h1 {
|
| 34 |
+
color: #4169e1;
|
| 35 |
+
text-align: center;
|
| 36 |
+
}
|
| 37 |
+
button {
|
| 38 |
+
background-color: #4169e1;
|
| 39 |
+
color: white;
|
| 40 |
+
border: none;
|
| 41 |
+
padding: 10px 15px;
|
| 42 |
+
border-radius: 4px;
|
| 43 |
+
cursor: pointer;
|
| 44 |
+
font-size: 16px;
|
| 45 |
+
margin: 10px 0;
|
| 46 |
+
}
|
| 47 |
+
button:hover {
|
| 48 |
+
background-color: #3158d3;
|
| 49 |
+
}
|
| 50 |
+
#canvas {
|
| 51 |
+
border: 1px solid #ddd;
|
| 52 |
+
display: block;
|
| 53 |
+
margin: 20px auto;
|
| 54 |
+
}
|
| 55 |
+
</style>
|
| 56 |
+
</head>
|
| 57 |
+
<body>
|
| 58 |
+
<div class="container">
|
| 59 |
+
<h1>Gradio HTML Test</h1>
|
| 60 |
+
<p>This is a simple HTML test page to verify that HTML rendering works in your Gradio application.</p>
|
| 61 |
+
|
| 62 |
+
<h2>Interactive Elements</h2>
|
| 63 |
+
<button id="testButton">Click Me!</button>
|
| 64 |
+
<p id="result">Button not clicked yet.</p>
|
| 65 |
+
|
| 66 |
+
<h2>Canvas Test</h2>
|
| 67 |
+
<canvas id="canvas" width="300" height="200"></canvas>
|
| 68 |
+
|
| 69 |
+
<h2>HTML Elements</h2>
|
| 70 |
+
<ul>
|
| 71 |
+
<li>List item 1</li>
|
| 72 |
+
<li>List item 2</li>
|
| 73 |
+
<li>List item 3</li>
|
| 74 |
+
</ul>
|
| 75 |
+
</div>
|
| 76 |
+
|
| 77 |
+
<script>
|
| 78 |
+
// Button click event
|
| 79 |
+
document.getElementById('testButton').addEventListener('click', function() {
|
| 80 |
+
document.getElementById('result').textContent = 'Button clicked at: ' + new Date().toLocaleTimeString();
|
| 81 |
+
});
|
| 82 |
+
|
| 83 |
+
// Canvas drawing
|
| 84 |
+
const canvas = document.getElementById('canvas');
|
| 85 |
+
const ctx = canvas.getContext('2d');
|
| 86 |
+
|
| 87 |
+
// Draw a simple scene
|
| 88 |
+
ctx.fillStyle = '#e0f0ff';
|
| 89 |
+
ctx.fillRect(0, 0, canvas.width, canvas.height);
|
| 90 |
+
|
| 91 |
+
// Draw a circle
|
| 92 |
+
ctx.beginPath();
|
| 93 |
+
ctx.arc(150, 100, 50, 0, Math.PI * 2);
|
| 94 |
+
ctx.fillStyle = '#4169e1';
|
| 95 |
+
ctx.fill();
|
| 96 |
+
|
| 97 |
+
// Draw a rectangle
|
| 98 |
+
ctx.fillStyle = '#ff6347';
|
| 99 |
+
ctx.fillRect(50, 50, 40, 40);
|
| 100 |
+
</script>
|
| 101 |
+
</body>
|
| 102 |
+
</html>
|
| 103 |
+
"""
|
| 104 |
+
|
| 105 |
+
game_html = """
|
| 106 |
+
<iframe id="gameFrame" style="width:660px; height:520px; border:none; display:block; margin:0 auto;" srcdoc='
|
| 107 |
+
<!DOCTYPE html>
|
| 108 |
+
<html>
|
| 109 |
+
<head>
|
| 110 |
+
<style>
|
| 111 |
+
body {
|
| 112 |
+
margin: 0;
|
| 113 |
+
padding: 0;
|
| 114 |
+
overflow: hidden;
|
| 115 |
+
font-family: sans-serif;
|
| 116 |
+
background: #1e1e1e;
|
| 117 |
+
}
|
| 118 |
+
#gameCanvas {
|
| 119 |
+
background: #70c5ce; /* sky-like */
|
| 120 |
+
display: block;
|
| 121 |
+
margin: 0 auto;
|
| 122 |
+
border: 2px solid black;
|
| 123 |
+
}
|
| 124 |
+
</style>
|
| 125 |
+
</head>
|
| 126 |
+
<body>
|
| 127 |
+
<canvas id="gameCanvas" width="640" height="480"></canvas>
|
| 128 |
+
<script>
|
| 129 |
+
// --- Simple "Red Baron"-style Game in Plain JS ---
|
| 130 |
+
|
| 131 |
+
// Grab the canvas and its context
|
| 132 |
+
const canvas = document.getElementById("gameCanvas");
|
| 133 |
+
const ctx = canvas.getContext("2d");
|
| 134 |
+
|
| 135 |
+
// Plane properties
|
| 136 |
+
let planeX = 50;
|
| 137 |
+
let planeY = canvas.height / 2;
|
| 138 |
+
const planeWidth = 40;
|
| 139 |
+
const planeHeight = 20;
|
| 140 |
+
const planeSpeed = 4;
|
| 141 |
+
|
| 142 |
+
// Bullet properties
|
| 143 |
+
let bullets = [];
|
| 144 |
+
const bulletSpeed = 6;
|
| 145 |
+
const bulletWidth = 6;
|
| 146 |
+
const bulletHeight = 2;
|
| 147 |
+
|
| 148 |
+
// Enemy properties
|
| 149 |
+
let enemies = [];
|
| 150 |
+
const enemyWidth = 40;
|
| 151 |
+
const enemyHeight = 20;
|
| 152 |
+
const enemySpeed = 2;
|
| 153 |
+
const spawnInterval = 100; // frames between spawns
|
| 154 |
+
let spawnCounter = 0;
|
| 155 |
+
|
| 156 |
+
// Key states
|
| 157 |
+
let keys = {
|
| 158 |
+
ArrowUp: false,
|
| 159 |
+
ArrowDown: false,
|
| 160 |
+
ArrowLeft: false,
|
| 161 |
+
ArrowRight: false,
|
| 162 |
+
Space: false
|
| 163 |
+
};
|
| 164 |
+
|
| 165 |
+
// Listen for keydown / keyup
|
| 166 |
+
document.addEventListener("keydown", (e) => {
|
| 167 |
+
if (keys.hasOwnProperty(e.code)) {
|
| 168 |
+
keys[e.code] = true;
|
| 169 |
+
}
|
| 170 |
+
});
|
| 171 |
+
document.addEventListener("keyup", (e) => {
|
| 172 |
+
if (keys.hasOwnProperty(e.code)) {
|
| 173 |
+
keys[e.code] = false;
|
| 174 |
+
}
|
| 175 |
+
});
|
| 176 |
+
|
| 177 |
+
// Main update loop
|
| 178 |
+
function update() {
|
| 179 |
+
// Move plane
|
| 180 |
+
if (keys["ArrowUp"] && planeY > 0) {
|
| 181 |
+
planeY -= planeSpeed;
|
| 182 |
+
}
|
| 183 |
+
if (keys["ArrowDown"] && planeY + planeHeight < canvas.height) {
|
| 184 |
+
planeY += planeSpeed;
|
| 185 |
+
}
|
| 186 |
+
if (keys["ArrowLeft"] && planeX > 0) {
|
| 187 |
+
planeX -= planeSpeed;
|
| 188 |
+
}
|
| 189 |
+
if (keys["ArrowRight"] && planeX + planeWidth < canvas.width) {
|
| 190 |
+
planeX += planeSpeed;
|
| 191 |
+
}
|
| 192 |
+
|
| 193 |
+
// Fire bullet (on every frame while space is held)
|
| 194 |
+
if (keys["Space"]) {
|
| 195 |
+
bullets.push({
|
| 196 |
+
x: planeX + planeWidth,
|
| 197 |
+
y: planeY + planeHeight / 2 - bulletHeight / 2,
|
| 198 |
+
w: bulletWidth,
|
| 199 |
+
h: bulletHeight
|
| 200 |
+
});
|
| 201 |
+
}
|
| 202 |
+
|
| 203 |
+
// Update bullets
|
| 204 |
+
for (let i = 0; i < bullets.length; i++) {
|
| 205 |
+
bullets[i].x += bulletSpeed;
|
| 206 |
+
}
|
| 207 |
+
// Remove bullets offscreen
|
| 208 |
+
bullets = bullets.filter((b) => b.x < canvas.width + 20);
|
| 209 |
+
|
| 210 |
+
// Spawn enemies periodically
|
| 211 |
+
spawnCounter++;
|
| 212 |
+
if (spawnCounter > spawnInterval) {
|
| 213 |
+
spawnCounter = 0;
|
| 214 |
+
enemies.push({
|
| 215 |
+
x: canvas.width,
|
| 216 |
+
y: Math.random() * (canvas.height - enemyHeight),
|
| 217 |
+
w: enemyWidth,
|
| 218 |
+
h: enemyHeight
|
| 219 |
+
});
|
| 220 |
+
}
|
| 221 |
+
|
| 222 |
+
// Update enemies
|
| 223 |
+
for (let i = 0; i < enemies.length; i++) {
|
| 224 |
+
enemies[i].x -= enemySpeed;
|
| 225 |
+
}
|
| 226 |
+
// Remove enemies offscreen
|
| 227 |
+
enemies = enemies.filter((e) => e.x > -enemyWidth);
|
| 228 |
+
|
| 229 |
+
// Check collisions bullets vs enemies
|
| 230 |
+
for (let e = enemies.length - 1; e >= 0; e--) {
|
| 231 |
+
let enemy = enemies[e];
|
| 232 |
+
for (let b = bullets.length - 1; b >= 0; b--) {
|
| 233 |
+
let bullet = bullets[b];
|
| 234 |
+
if (
|
| 235 |
+
bullet.x < enemy.x + enemy.w &&
|
| 236 |
+
bullet.x + bullet.w > enemy.x &&
|
| 237 |
+
bullet.y < enemy.y + enemy.h &&
|
| 238 |
+
bullet.y + bullet.h > enemy.y
|
| 239 |
+
) {
|
| 240 |
+
// collision
|
| 241 |
+
enemies.splice(e, 1);
|
| 242 |
+
bullets.splice(b, 1);
|
| 243 |
+
break;
|
| 244 |
+
}
|
| 245 |
+
}
|
| 246 |
+
}
|
| 247 |
+
|
| 248 |
+
// Draw everything
|
| 249 |
+
draw();
|
| 250 |
+
requestAnimationFrame(update);
|
| 251 |
+
}
|
| 252 |
+
|
| 253 |
+
// Render the scene
|
| 254 |
+
function draw() {
|
| 255 |
+
// Clear
|
| 256 |
+
ctx.clearRect(0, 0, canvas.width, canvas.height);
|
| 257 |
+
|
| 258 |
+
// Draw plane (red rectangle for demonstration)
|
| 259 |
+
ctx.fillStyle = "red";
|
| 260 |
+
ctx.fillRect(planeX, planeY, planeWidth, planeHeight);
|
| 261 |
+
|
| 262 |
+
// Draw bullets (small black rectangles)
|
| 263 |
+
ctx.fillStyle = "black";
|
| 264 |
+
bullets.forEach((b) => {
|
| 265 |
+
ctx.fillRect(b.x, b.y, b.w, b.h);
|
| 266 |
+
});
|
| 267 |
+
|
| 268 |
+
// Draw enemies (simple gray rectangles)
|
| 269 |
+
ctx.fillStyle = "gray";
|
| 270 |
+
enemies.forEach((en) => {
|
| 271 |
+
ctx.fillRect(en.x, en.y, en.w, en.h);
|
| 272 |
+
});
|
| 273 |
+
}
|
| 274 |
+
|
| 275 |
+
// Start game loop
|
| 276 |
+
window.onload = function() {
|
| 277 |
+
// Make sure canvas is fully loaded
|
| 278 |
+
update();
|
| 279 |
+
};
|
| 280 |
+
</script>
|
| 281 |
+
</body>
|
| 282 |
+
</html>
|
| 283 |
+
'></iframe>
|
| 284 |
+
<div style="text-align:center; margin-top:10px;">
|
| 285 |
+
<p>Controls: Arrow keys to move, Space to shoot</p>
|
| 286 |
+
</div>
|
| 287 |
+
"""
|
| 288 |
+
|
| 289 |
+
#######################################################
|
| 290 |
+
# Dictionary of game -> directory paths
|
| 291 |
+
#######################################################
|
| 292 |
+
GAMES = {
|
| 293 |
+
"Super Mario Bros": "assets/super_mario_bros",
|
| 294 |
+
"Sokoban": "assets/sokoban",
|
| 295 |
+
"Tetris": "assets/tetris",
|
| 296 |
+
"2048": "assets/2048",
|
| 297 |
+
"Candy Crash": "assets/candy"
|
| 298 |
+
}
|
| 299 |
+
# Ensure each directory exists
|
| 300 |
+
for path in GAMES.values():
|
| 301 |
+
os.makedirs(path, exist_ok=True)
|
| 302 |
+
|
| 303 |
+
#######################################################
|
| 304 |
+
# Scoreboard data for each game
|
| 305 |
+
#######################################################
|
| 306 |
+
|
| 307 |
+
# TODO (lanxiang): read actual data here
|
| 308 |
+
mario_scores = [
|
| 309 |
+
["Alice", 9000, "3/5", "00:35"],
|
| 310 |
+
["Bob", 2500, "2/5", "00:12"],
|
| 311 |
+
["Carol", 500, "1/5", "00:05"]
|
| 312 |
+
]
|
| 313 |
+
sokoban_scores = [
|
| 314 |
+
["Alice", 100, 3],
|
| 315 |
+
["Bob", 350, 2],
|
| 316 |
+
["Carol", 500, 1]
|
| 317 |
+
]
|
| 318 |
+
|
| 319 |
+
tetris_scores = [
|
| 320 |
+
["Alice", 15000, 120],
|
| 321 |
+
["Bob", 8000, 60],
|
| 322 |
+
["Carol", 4000, 45]
|
| 323 |
+
]
|
| 324 |
+
|
| 325 |
+
candy_scores = [
|
| 326 |
+
["Alice", 12000, 10],
|
| 327 |
+
["Bob", 9500, 8],
|
| 328 |
+
["Carol", 3000, 5]
|
| 329 |
+
]
|
| 330 |
+
|
| 331 |
+
# 2048 columns: [Player, Scores, #Steps]
|
| 332 |
+
game_2048_scores = [
|
| 333 |
+
["Alice", 8192, 300],
|
| 334 |
+
["Bob", 4096, 200],
|
| 335 |
+
["Carol", 2048, 150]
|
| 336 |
+
]
|
| 337 |
+
|
| 338 |
+
#######################################################
|
| 339 |
+
# Functions to return the scoreboard as DataFrames
|
| 340 |
+
#######################################################
|
| 341 |
+
def get_mario_leaderboard():
|
| 342 |
+
return pd.DataFrame(
|
| 343 |
+
mario_scores,
|
| 344 |
+
columns=["Player", "Progress (current/total)", "Score", "Time"]
|
| 345 |
+
)
|
| 346 |
+
|
| 347 |
+
def get_sokoban_leaderboard():
|
| 348 |
+
return pd.DataFrame(
|
| 349 |
+
sokoban_scores,
|
| 350 |
+
columns=["Player", "Levels Cracked", "Steps"]
|
| 351 |
+
)
|
| 352 |
+
|
| 353 |
+
def get_tetris_leaderboard():
|
| 354 |
+
return pd.DataFrame(
|
| 355 |
+
tetris_scores,
|
| 356 |
+
columns=["Player", "Scores", "Steps"]
|
| 357 |
+
)
|
| 358 |
+
|
| 359 |
+
def get_candy_leaderboard():
|
| 360 |
+
return pd.DataFrame(
|
| 361 |
+
candy_scores,
|
| 362 |
+
columns=["Player", "Levels Cracked", "Scores"]
|
| 363 |
+
)
|
| 364 |
+
|
| 365 |
+
def get_2048_leaderboard():
|
| 366 |
+
return pd.DataFrame(
|
| 367 |
+
game_2048_scores,
|
| 368 |
+
columns=["Player", "Scores", "Steps"]
|
| 369 |
+
)
|
| 370 |
+
|
| 371 |
+
#######################################################
|
| 372 |
+
# GIF Handling
|
| 373 |
+
#######################################################
|
| 374 |
+
def create_or_update_resized_gif(original_path, max_dim=600):
|
| 375 |
+
base, ext = os.path.splitext(original_path)
|
| 376 |
+
resized_path = f"{base}_resized{ext}"
|
| 377 |
+
if os.path.exists(resized_path):
|
| 378 |
+
return resized_path
|
| 379 |
+
|
| 380 |
+
with Image.open(original_path) as im:
|
| 381 |
+
w, h = im.size
|
| 382 |
+
needs_resize = (w > max_dim or h > max_dim)
|
| 383 |
+
|
| 384 |
+
frames = []
|
| 385 |
+
for frame in ImageSequence.Iterator(im):
|
| 386 |
+
frame_rgba = frame.convert("RGBA")
|
| 387 |
+
if needs_resize:
|
| 388 |
+
ratio = min(max_dim / w, max_dim / h)
|
| 389 |
+
new_w, new_h = int(w * ratio), int(h * ratio)
|
| 390 |
+
frame_rgba = frame_rgba.resize((new_w, new_h), Image.LANCZOS)
|
| 391 |
+
frames.append(frame_rgba.convert("P"))
|
| 392 |
+
|
| 393 |
+
output_bytes = io.BytesIO()
|
| 394 |
+
frames[0].save(
|
| 395 |
+
output_bytes,
|
| 396 |
+
format="GIF",
|
| 397 |
+
save_all=True,
|
| 398 |
+
append_images=frames[1:],
|
| 399 |
+
loop=0,
|
| 400 |
+
disposal=2,
|
| 401 |
+
optimize=False
|
| 402 |
+
)
|
| 403 |
+
output_bytes.seek(0)
|
| 404 |
+
|
| 405 |
+
with open(resized_path, "wb") as f_out:
|
| 406 |
+
f_out.write(output_bytes.read())
|
| 407 |
+
return resized_path
|
| 408 |
+
|
| 409 |
+
def list_gifs(game_name):
|
| 410 |
+
gif_dir = GAMES[game_name]
|
| 411 |
+
all_gifs = [
|
| 412 |
+
os.path.join(gif_dir, f)
|
| 413 |
+
for f in os.listdir(gif_dir)
|
| 414 |
+
if f.lower().endswith(".gif") and not f.lower().endswith("_resized.gif")
|
| 415 |
+
]
|
| 416 |
+
resized_paths = []
|
| 417 |
+
for gif_path in all_gifs:
|
| 418 |
+
resized_gif_path = create_or_update_resized_gif(gif_path, max_dim=600)
|
| 419 |
+
resized_paths.append(resized_gif_path)
|
| 420 |
+
|
| 421 |
+
previously_resized = [
|
| 422 |
+
os.path.join(gif_dir, f)
|
| 423 |
+
for f in os.listdir(gif_dir)
|
| 424 |
+
if f.lower().endswith("_resized.gif")
|
| 425 |
+
]
|
| 426 |
+
all_resized = list(set(resized_paths + previously_resized))
|
| 427 |
+
return sorted(all_resized)
|
| 428 |
+
|
| 429 |
+
#######################################################
|
| 430 |
+
# Custom CSS
|
| 431 |
+
#######################################################
|
| 432 |
+
fancy_css = """
|
| 433 |
+
body {
|
| 434 |
+
font-family: 'Trebuchet MS', sans-serif;
|
| 435 |
+
background: #f0f8ff;
|
| 436 |
+
color: #333333;
|
| 437 |
+
}
|
| 438 |
+
h1 {
|
| 439 |
+
color: #4b9cd3;
|
| 440 |
+
text-align: center;
|
| 441 |
+
margin-top: 20px;
|
| 442 |
+
}
|
| 443 |
+
.gradio-container {
|
| 444 |
+
max-width: 800px;
|
| 445 |
+
margin: 0 auto;
|
| 446 |
+
padding: 20px;
|
| 447 |
+
}
|
| 448 |
+
"""
|
| 449 |
+
|
| 450 |
+
#######################################################
|
| 451 |
+
# Build the App
|
| 452 |
+
#######################################################
|
| 453 |
+
def build_app():
|
| 454 |
+
with gr.Blocks(css=fancy_css) as demo:
|
| 455 |
+
gr.Markdown("# Game Arena: Gaming Agent")
|
| 456 |
+
|
| 457 |
+
with gr.Tabs():
|
| 458 |
+
# tab: "Gallery"
|
| 459 |
+
with gr.Tab("Gallery"):
|
| 460 |
+
with gr.Tabs():
|
| 461 |
+
for game_name in GAMES:
|
| 462 |
+
with gr.Tab(game_name):
|
| 463 |
+
gr.Markdown(f"### {game_name} Gallery")
|
| 464 |
+
gr.Gallery(
|
| 465 |
+
label="GIFs",
|
| 466 |
+
value=list_gifs(game_name)
|
| 467 |
+
)
|
| 468 |
+
|
| 469 |
+
# tab: "Leaderboard"
|
| 470 |
+
with gr.Tab("Leaderboard"):
|
| 471 |
+
gr.Markdown("## Game Leaderboards")
|
| 472 |
+
|
| 473 |
+
# Sub-tabs for each game
|
| 474 |
+
with gr.Tabs():
|
| 475 |
+
with gr.Tab("Mario"):
|
| 476 |
+
gr.Markdown("### Mario Leaderboard")
|
| 477 |
+
gr.DataFrame(value=get_mario_leaderboard(), interactive=False)
|
| 478 |
+
|
| 479 |
+
with gr.Tab("Sokoban"):
|
| 480 |
+
gr.Markdown("### Sokoban Leaderboard")
|
| 481 |
+
gr.DataFrame(value=get_sokoban_leaderboard(), interactive=False)
|
| 482 |
+
|
| 483 |
+
with gr.Tab("Tetris"):
|
| 484 |
+
gr.Markdown("### Tetris Leaderboard")
|
| 485 |
+
gr.DataFrame(value=get_tetris_leaderboard(), interactive=False)
|
| 486 |
+
|
| 487 |
+
with gr.Tab("2048"):
|
| 488 |
+
gr.Markdown("### 2048 Leaderboard")
|
| 489 |
+
gr.DataFrame(value=get_2048_leaderboard(), interactive=False)
|
| 490 |
+
|
| 491 |
+
with gr.Tab("Candy Crash"):
|
| 492 |
+
gr.Markdown("### Candy Crash Leaderboard")
|
| 493 |
+
gr.DataFrame(value=get_candy_leaderboard(), interactive=False)
|
| 494 |
+
|
| 495 |
+
# Top-level tab: "Red Baron" game demo
|
| 496 |
+
with gr.Tab("Red Baron"):
|
| 497 |
+
gr.Markdown("## Red Baron Game Demo")
|
| 498 |
+
gr.HTML(game_html)
|
| 499 |
+
|
| 500 |
+
return demo
|
| 501 |
+
|
| 502 |
+
if __name__ == "__main__":
|
| 503 |
+
demo_app = build_app()
|
| 504 |
+
# demo_app.launch(server_name="0.0.0.0", server_port=7860)
|
| 505 |
+
demo_app.launch(server_name="127.0.0.1", server_port=7860, debug=True)
|
gradio_app_v2.py
ADDED
|
@@ -0,0 +1,540 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import os
|
| 3 |
+
import pandas as pd
|
| 4 |
+
import json
|
| 5 |
+
from PIL import Image, ImageSequence
|
| 6 |
+
import io
|
| 7 |
+
from functools import reduce
|
| 8 |
+
import numpy as np
|
| 9 |
+
from datetime import datetime, timedelta
|
| 10 |
+
import matplotlib.pyplot as plt
|
| 11 |
+
from leaderboard_utils import (
|
| 12 |
+
get_organization,
|
| 13 |
+
get_mario_leaderboard,
|
| 14 |
+
get_sokoban_leaderboard,
|
| 15 |
+
get_2048_leaderboard,
|
| 16 |
+
get_candy_leaderboard,
|
| 17 |
+
get_tetris_leaderboard,
|
| 18 |
+
get_tetris_planning_leaderboard,
|
| 19 |
+
get_combined_leaderboard,
|
| 20 |
+
GAME_ORDER
|
| 21 |
+
)
|
| 22 |
+
from data_visualization import (
|
| 23 |
+
get_combined_leaderboard_with_radar,
|
| 24 |
+
create_organization_radar_chart,
|
| 25 |
+
create_top_players_radar_chart,
|
| 26 |
+
create_player_radar_chart,
|
| 27 |
+
create_horizontal_bar_chart
|
| 28 |
+
)
|
| 29 |
+
|
| 30 |
+
# Define time points and their corresponding data files
|
| 31 |
+
TIME_POINTS = {
|
| 32 |
+
"03/25/2025": "rank_data_03_25_2025.json",
|
| 33 |
+
# Add more time points here as they become available
|
| 34 |
+
}
|
| 35 |
+
|
| 36 |
+
# Load the initial JSON file with rank data
|
| 37 |
+
with open(TIME_POINTS["03/25/2025"], "r") as f:
|
| 38 |
+
rank_data = json.load(f)
|
| 39 |
+
|
| 40 |
+
# Add leaderboard state at the top level
|
| 41 |
+
leaderboard_state = {
|
| 42 |
+
"current_game": None,
|
| 43 |
+
"previous_overall": {
|
| 44 |
+
"Super Mario Bros": True,
|
| 45 |
+
"Sokoban": True,
|
| 46 |
+
"2048": True,
|
| 47 |
+
"Candy Crash": True,
|
| 48 |
+
"Tetris (complete)": True,
|
| 49 |
+
"Tetris (planning only)": True
|
| 50 |
+
},
|
| 51 |
+
"previous_details": {
|
| 52 |
+
"Super Mario Bros": False,
|
| 53 |
+
"Sokoban": False,
|
| 54 |
+
"2048": False,
|
| 55 |
+
"Candy Crash": False,
|
| 56 |
+
"Tetris (complete)": False,
|
| 57 |
+
"Tetris (planning only)": False
|
| 58 |
+
}
|
| 59 |
+
}
|
| 60 |
+
|
| 61 |
+
# Define GIF paths for the carousel
|
| 62 |
+
GIF_PATHS = [
|
| 63 |
+
"assets/super_mario_bros/super_mario.gif",
|
| 64 |
+
"assets/sokoban/sokoban.gif",
|
| 65 |
+
"assets/2048/2048.gif",
|
| 66 |
+
"assets/candy/candy.gif",
|
| 67 |
+
"assets/tetris/tetris.gif"
|
| 68 |
+
]
|
| 69 |
+
|
| 70 |
+
# Print and verify GIF paths
|
| 71 |
+
print("\nChecking GIF paths:")
|
| 72 |
+
for gif_path in GIF_PATHS:
|
| 73 |
+
if os.path.exists(gif_path):
|
| 74 |
+
print(f"✓ Found: {gif_path}")
|
| 75 |
+
# Print file size
|
| 76 |
+
size = os.path.getsize(gif_path)
|
| 77 |
+
print(f" Size: {size / (1024*1024):.2f} MB")
|
| 78 |
+
else:
|
| 79 |
+
print(f"✗ Missing: {gif_path}")
|
| 80 |
+
|
| 81 |
+
def load_gif(gif_path):
|
| 82 |
+
"""Load a GIF file and return it as a PIL Image"""
|
| 83 |
+
try:
|
| 84 |
+
img = Image.open(gif_path)
|
| 85 |
+
print(f"Successfully loaded GIF: {gif_path}")
|
| 86 |
+
return img
|
| 87 |
+
except Exception as e:
|
| 88 |
+
print(f"Error loading GIF {gif_path}: {e}")
|
| 89 |
+
return None
|
| 90 |
+
|
| 91 |
+
def create_gif_carousel():
|
| 92 |
+
"""Create a custom HTML/JS component for GIF carousel"""
|
| 93 |
+
print("\nCreating GIF carousel with paths:", GIF_PATHS)
|
| 94 |
+
html = f"""
|
| 95 |
+
<div id="gif-carousel" style="width: 100%; height: 300px; position: relative; background-color: #f0f0f0;">
|
| 96 |
+
<img id="current-gif" style="width: 100%; height: 100%; object-fit: contain;" onerror="console.error('Failed to load GIF:', this.src);">
|
| 97 |
+
</div>
|
| 98 |
+
<script>
|
| 99 |
+
const gifs = {json.dumps(GIF_PATHS)};
|
| 100 |
+
let currentIndex = 0;
|
| 101 |
+
|
| 102 |
+
function updateGif() {{
|
| 103 |
+
const img = document.getElementById('current-gif');
|
| 104 |
+
console.log('Loading GIF:', gifs[currentIndex]);
|
| 105 |
+
img.src = gifs[currentIndex];
|
| 106 |
+
currentIndex = (currentIndex + 1) % gifs.length;
|
| 107 |
+
}}
|
| 108 |
+
|
| 109 |
+
// Update GIF every 5 seconds
|
| 110 |
+
setInterval(updateGif, 5000);
|
| 111 |
+
// Initial load
|
| 112 |
+
updateGif();
|
| 113 |
+
</script>
|
| 114 |
+
"""
|
| 115 |
+
return gr.HTML(html)
|
| 116 |
+
|
| 117 |
+
def load_rank_data(time_point):
|
| 118 |
+
"""Load rank data for a specific time point"""
|
| 119 |
+
if time_point in TIME_POINTS:
|
| 120 |
+
try:
|
| 121 |
+
with open(TIME_POINTS[time_point], "r") as f:
|
| 122 |
+
return json.load(f)
|
| 123 |
+
except FileNotFoundError:
|
| 124 |
+
return None
|
| 125 |
+
return None
|
| 126 |
+
|
| 127 |
+
def update_leaderboard(mario_overall, mario_details,
|
| 128 |
+
sokoban_overall, sokoban_details,
|
| 129 |
+
_2048_overall, _2048_details,
|
| 130 |
+
candy_overall, candy_details,
|
| 131 |
+
tetris_overall, tetris_details,
|
| 132 |
+
tetris_plan_overall, tetris_plan_details):
|
| 133 |
+
global leaderboard_state
|
| 134 |
+
|
| 135 |
+
# Convert current checkbox states to dictionary for easier comparison
|
| 136 |
+
current_overall = {
|
| 137 |
+
"Super Mario Bros": mario_overall,
|
| 138 |
+
"Sokoban": sokoban_overall,
|
| 139 |
+
"2048": _2048_overall,
|
| 140 |
+
"Candy Crash": candy_overall,
|
| 141 |
+
"Tetris (complete)": tetris_overall,
|
| 142 |
+
"Tetris (planning only)": tetris_plan_overall
|
| 143 |
+
}
|
| 144 |
+
|
| 145 |
+
current_details = {
|
| 146 |
+
"Super Mario Bros": mario_details,
|
| 147 |
+
"Sokoban": sokoban_details,
|
| 148 |
+
"2048": _2048_details,
|
| 149 |
+
"Candy Crash": candy_details,
|
| 150 |
+
"Tetris (complete)": tetris_details,
|
| 151 |
+
"Tetris (planning only)": tetris_plan_details
|
| 152 |
+
}
|
| 153 |
+
|
| 154 |
+
# Find which game's state changed
|
| 155 |
+
changed_game = None
|
| 156 |
+
for game in current_overall.keys():
|
| 157 |
+
if (current_overall[game] != leaderboard_state["previous_overall"][game] or
|
| 158 |
+
current_details[game] != leaderboard_state["previous_details"][game]):
|
| 159 |
+
changed_game = game
|
| 160 |
+
break
|
| 161 |
+
|
| 162 |
+
if changed_game:
|
| 163 |
+
# If a game's details checkbox was checked
|
| 164 |
+
if current_details[changed_game] and not leaderboard_state["previous_details"][changed_game]:
|
| 165 |
+
# Reset all other games' states
|
| 166 |
+
for game in current_overall.keys():
|
| 167 |
+
if game != changed_game:
|
| 168 |
+
current_overall[game] = False
|
| 169 |
+
current_details[game] = False
|
| 170 |
+
leaderboard_state["previous_overall"][game] = False
|
| 171 |
+
leaderboard_state["previous_details"][game] = False
|
| 172 |
+
|
| 173 |
+
# Update state for the selected game
|
| 174 |
+
leaderboard_state["current_game"] = changed_game
|
| 175 |
+
leaderboard_state["previous_overall"][changed_game] = True # Set overall to True when details is checked
|
| 176 |
+
leaderboard_state["previous_details"][changed_game] = True
|
| 177 |
+
current_overall[changed_game] = True # Ensure the overall checkbox is checked
|
| 178 |
+
|
| 179 |
+
# If a game's overall checkbox was checked
|
| 180 |
+
elif current_overall[changed_game] and not leaderboard_state["previous_overall"][changed_game]:
|
| 181 |
+
# If we were in details view for another game, switch to overall view
|
| 182 |
+
if leaderboard_state["current_game"] and leaderboard_state["previous_details"][leaderboard_state["current_game"]]:
|
| 183 |
+
# Reset previous game's details
|
| 184 |
+
leaderboard_state["previous_details"][leaderboard_state["current_game"]] = False
|
| 185 |
+
current_details[leaderboard_state["current_game"]] = False
|
| 186 |
+
leaderboard_state["current_game"] = None
|
| 187 |
+
|
| 188 |
+
# Update state
|
| 189 |
+
leaderboard_state["previous_overall"][changed_game] = True
|
| 190 |
+
leaderboard_state["previous_details"][changed_game] = False
|
| 191 |
+
|
| 192 |
+
# If a game's overall checkbox was unchecked
|
| 193 |
+
elif not current_overall[changed_game] and leaderboard_state["previous_overall"][changed_game]:
|
| 194 |
+
# If we're in details view, don't allow unchecking the overall checkbox
|
| 195 |
+
if leaderboard_state["current_game"] == changed_game:
|
| 196 |
+
current_overall[changed_game] = True
|
| 197 |
+
else:
|
| 198 |
+
leaderboard_state["previous_overall"][changed_game] = False
|
| 199 |
+
if leaderboard_state["current_game"] == changed_game:
|
| 200 |
+
leaderboard_state["current_game"] = None
|
| 201 |
+
|
| 202 |
+
# If a game's details checkbox was unchecked
|
| 203 |
+
elif not current_details[changed_game] and leaderboard_state["previous_details"][changed_game]:
|
| 204 |
+
leaderboard_state["previous_details"][changed_game] = False
|
| 205 |
+
if leaderboard_state["current_game"] == changed_game:
|
| 206 |
+
leaderboard_state["current_game"] = None
|
| 207 |
+
|
| 208 |
+
# Build dictionary for selected games
|
| 209 |
+
selected_games = {
|
| 210 |
+
"Super Mario Bros": current_overall["Super Mario Bros"],
|
| 211 |
+
"Sokoban": current_overall["Sokoban"],
|
| 212 |
+
"2048": current_overall["2048"],
|
| 213 |
+
"Candy Crash": current_overall["Candy Crash"],
|
| 214 |
+
"Tetris (complete)": current_overall["Tetris (complete)"],
|
| 215 |
+
"Tetris (planning only)": current_overall["Tetris (planning only)"]
|
| 216 |
+
}
|
| 217 |
+
|
| 218 |
+
# Filter GIF paths based on selected games
|
| 219 |
+
filtered_gifs = []
|
| 220 |
+
if current_overall["Super Mario Bros"]:
|
| 221 |
+
filtered_gifs.append(GIF_PATHS[0])
|
| 222 |
+
if current_overall["Sokoban"]:
|
| 223 |
+
filtered_gifs.append(GIF_PATHS[1])
|
| 224 |
+
if current_overall["2048"]:
|
| 225 |
+
filtered_gifs.append(GIF_PATHS[2])
|
| 226 |
+
if current_overall["Candy Crash"]:
|
| 227 |
+
filtered_gifs.append(GIF_PATHS[3])
|
| 228 |
+
if current_overall["Tetris (complete)"] or current_overall["Tetris (planning only)"]:
|
| 229 |
+
filtered_gifs.append(GIF_PATHS[4])
|
| 230 |
+
|
| 231 |
+
# Get the appropriate DataFrame and chart based on current state
|
| 232 |
+
if leaderboard_state["current_game"]:
|
| 233 |
+
# For detailed view
|
| 234 |
+
if leaderboard_state["current_game"] == "Super Mario Bros":
|
| 235 |
+
df = get_mario_leaderboard(rank_data)
|
| 236 |
+
elif leaderboard_state["current_game"] == "Sokoban":
|
| 237 |
+
df = get_sokoban_leaderboard(rank_data)
|
| 238 |
+
elif leaderboard_state["current_game"] == "2048":
|
| 239 |
+
df = get_2048_leaderboard(rank_data)
|
| 240 |
+
elif leaderboard_state["current_game"] == "Candy Crash":
|
| 241 |
+
df = get_candy_leaderboard(rank_data)
|
| 242 |
+
elif leaderboard_state["current_game"] == "Tetris (complete)":
|
| 243 |
+
df = get_tetris_leaderboard(rank_data)
|
| 244 |
+
else: # Tetris (planning only)
|
| 245 |
+
df = get_tetris_planning_leaderboard(rank_data)
|
| 246 |
+
|
| 247 |
+
# Always create a new chart for detailed view
|
| 248 |
+
chart = create_horizontal_bar_chart(df, leaderboard_state["current_game"])
|
| 249 |
+
else:
|
| 250 |
+
# For overall view
|
| 251 |
+
df = get_combined_leaderboard(rank_data, selected_games)
|
| 252 |
+
_, chart = get_combined_leaderboard_with_radar(rank_data, selected_games)
|
| 253 |
+
|
| 254 |
+
return (df, chart, filtered_gifs,
|
| 255 |
+
current_overall["Super Mario Bros"], current_details["Super Mario Bros"],
|
| 256 |
+
current_overall["Sokoban"], current_details["Sokoban"],
|
| 257 |
+
current_overall["2048"], current_details["2048"],
|
| 258 |
+
current_overall["Candy Crash"], current_details["Candy Crash"],
|
| 259 |
+
current_overall["Tetris (complete)"], current_details["Tetris (complete)"],
|
| 260 |
+
current_overall["Tetris (planning only)"], current_details["Tetris (planning only)"])
|
| 261 |
+
|
| 262 |
+
def update_leaderboard_with_time(time_point, mario_overall, mario_details,
|
| 263 |
+
sokoban_overall, sokoban_details,
|
| 264 |
+
_2048_overall, _2048_details,
|
| 265 |
+
candy_overall, candy_details,
|
| 266 |
+
tetris_overall, tetris_details,
|
| 267 |
+
tetris_plan_overall, tetris_plan_details):
|
| 268 |
+
# Load rank data for the selected time point
|
| 269 |
+
global rank_data
|
| 270 |
+
new_rank_data = load_rank_data(time_point)
|
| 271 |
+
if new_rank_data is not None:
|
| 272 |
+
rank_data = new_rank_data
|
| 273 |
+
|
| 274 |
+
# Use the existing update_leaderboard function
|
| 275 |
+
return update_leaderboard(mario_overall, mario_details,
|
| 276 |
+
sokoban_overall, sokoban_details,
|
| 277 |
+
_2048_overall, _2048_details,
|
| 278 |
+
candy_overall, candy_details,
|
| 279 |
+
tetris_overall, tetris_details,
|
| 280 |
+
tetris_plan_overall, tetris_plan_details)
|
| 281 |
+
|
| 282 |
+
def clear_filters():
|
| 283 |
+
global leaderboard_state
|
| 284 |
+
|
| 285 |
+
# Reset all checkboxes to default state and get fresh data
|
| 286 |
+
df = get_combined_leaderboard(rank_data, {
|
| 287 |
+
"Super Mario Bros": True,
|
| 288 |
+
"Sokoban": True,
|
| 289 |
+
"2048": True,
|
| 290 |
+
"Candy Crash": True,
|
| 291 |
+
"Tetris (complete)": True,
|
| 292 |
+
"Tetris (planning only)": True
|
| 293 |
+
})
|
| 294 |
+
|
| 295 |
+
# Get the radar chart visualization
|
| 296 |
+
_, chart = get_combined_leaderboard_with_radar(rank_data, {
|
| 297 |
+
"Super Mario Bros": True,
|
| 298 |
+
"Sokoban": True,
|
| 299 |
+
"2048": True,
|
| 300 |
+
"Candy Crash": True,
|
| 301 |
+
"Tetris (complete)": True,
|
| 302 |
+
"Tetris (planning only)": True
|
| 303 |
+
})
|
| 304 |
+
|
| 305 |
+
# Reset the leaderboard state to match the default checkbox states
|
| 306 |
+
leaderboard_state = {
|
| 307 |
+
"current_game": None,
|
| 308 |
+
"previous_overall": {
|
| 309 |
+
"Super Mario Bros": True,
|
| 310 |
+
"Sokoban": True,
|
| 311 |
+
"2048": True,
|
| 312 |
+
"Candy Crash": True,
|
| 313 |
+
"Tetris (complete)": True,
|
| 314 |
+
"Tetris (planning only)": True
|
| 315 |
+
},
|
| 316 |
+
"previous_details": {
|
| 317 |
+
"Super Mario Bros": False,
|
| 318 |
+
"Sokoban": False,
|
| 319 |
+
"2048": False,
|
| 320 |
+
"Candy Crash": False,
|
| 321 |
+
"Tetris (complete)": False,
|
| 322 |
+
"Tetris (planning only)": False
|
| 323 |
+
}
|
| 324 |
+
}
|
| 325 |
+
|
| 326 |
+
# Return both the DataFrame and the visualization
|
| 327 |
+
return (df, chart, GIF_PATHS,
|
| 328 |
+
True, False, # mario
|
| 329 |
+
True, False, # sokoban
|
| 330 |
+
True, False, # 2048
|
| 331 |
+
True, False, # candy
|
| 332 |
+
True, False, # tetris
|
| 333 |
+
True, False) # tetris plan
|
| 334 |
+
|
| 335 |
+
def build_app():
|
| 336 |
+
with gr.Blocks(css="""
|
| 337 |
+
.gallery-container {
|
| 338 |
+
height: 50vh !important;
|
| 339 |
+
max-height: 600px !important;
|
| 340 |
+
min-height: 300px !important;
|
| 341 |
+
background-color: #f8f9fa;
|
| 342 |
+
border-radius: 10px;
|
| 343 |
+
padding: 5px !important;
|
| 344 |
+
box-shadow: 0 2px 4px rgba(0,0,0,0.1);
|
| 345 |
+
overflow: hidden;
|
| 346 |
+
aspect-ratio: 1 !important;
|
| 347 |
+
display: flex !important;
|
| 348 |
+
align-items: center !important;
|
| 349 |
+
justify-content: center !important;
|
| 350 |
+
}
|
| 351 |
+
.gallery-container .gallery-item {
|
| 352 |
+
height: 100% !important;
|
| 353 |
+
width: 100% !important;
|
| 354 |
+
border-radius: 8px;
|
| 355 |
+
overflow: hidden;
|
| 356 |
+
margin: 0 !important;
|
| 357 |
+
padding: 0 !important;
|
| 358 |
+
aspect-ratio: 1 !important;
|
| 359 |
+
display: flex !important;
|
| 360 |
+
align-items: center !important;
|
| 361 |
+
justify-content: center !important;
|
| 362 |
+
}
|
| 363 |
+
.gallery-container .gallery-item img {
|
| 364 |
+
height: 100% !important;
|
| 365 |
+
width: 100% !important;
|
| 366 |
+
object-fit: contain !important;
|
| 367 |
+
aspect-ratio: 1 !important;
|
| 368 |
+
}
|
| 369 |
+
.visualization-container {
|
| 370 |
+
height: 50vh !important;
|
| 371 |
+
max-height: 600px !important;
|
| 372 |
+
min-height: 300px !important;
|
| 373 |
+
background-color: #f8f9fa;
|
| 374 |
+
border-radius: 10px;
|
| 375 |
+
padding: 15px;
|
| 376 |
+
box-shadow: 0 2px 4px rgba(0,0,0,0.1);
|
| 377 |
+
overflow: hidden;
|
| 378 |
+
margin-left: 10px !important; /* Add small gap between gallery and visualization */
|
| 379 |
+
}
|
| 380 |
+
.visualization-container .plot {
|
| 381 |
+
height: 100% !important;
|
| 382 |
+
width: 100% !important;
|
| 383 |
+
}
|
| 384 |
+
.section-title {
|
| 385 |
+
font-size: 1.5em;
|
| 386 |
+
font-weight: bold;
|
| 387 |
+
color: #2c3e50;
|
| 388 |
+
margin-bottom: 15px;
|
| 389 |
+
padding-bottom: 10px;
|
| 390 |
+
border-bottom: 2px solid #e9ecef;
|
| 391 |
+
}
|
| 392 |
+
/* Add container for the entire app */
|
| 393 |
+
.container {
|
| 394 |
+
max-width: 1400px;
|
| 395 |
+
margin: 0 auto;
|
| 396 |
+
padding: 0 20px;
|
| 397 |
+
}
|
| 398 |
+
/* Add flex layout for the row containing gallery and visualization */
|
| 399 |
+
.gallery-viz-row {
|
| 400 |
+
display: flex !important;
|
| 401 |
+
align-items: center !important;
|
| 402 |
+
gap: 20px !important; /* Add consistent gap between components */
|
| 403 |
+
}
|
| 404 |
+
""") as demo:
|
| 405 |
+
gr.Markdown("# 🎮 Game Arena: Gaming Agent 🎲")
|
| 406 |
+
|
| 407 |
+
with gr.Tabs():
|
| 408 |
+
with gr.Tab("🏆 Leaderboard"):
|
| 409 |
+
# Visualization section at the very top
|
| 410 |
+
with gr.Row():
|
| 411 |
+
gr.Markdown("### 📊 Gallery")
|
| 412 |
+
with gr.Row(elem_classes="gallery-viz-row"):
|
| 413 |
+
# Split into two columns
|
| 414 |
+
with gr.Column(scale=3):
|
| 415 |
+
gallery = gr.Gallery(
|
| 416 |
+
value=GIF_PATHS,
|
| 417 |
+
label="Game Demos",
|
| 418 |
+
show_label=True,
|
| 419 |
+
elem_id="gallery",
|
| 420 |
+
elem_classes="gallery-container",
|
| 421 |
+
columns=1,
|
| 422 |
+
rows=1,
|
| 423 |
+
min_width=100,
|
| 424 |
+
container=True,
|
| 425 |
+
allow_preview=True,
|
| 426 |
+
object_fit='contain',
|
| 427 |
+
show_download_button=False,
|
| 428 |
+
show_share_button=False,
|
| 429 |
+
show_fullscreen_button=True
|
| 430 |
+
)
|
| 431 |
+
with gr.Column(scale=4):
|
| 432 |
+
visualization = gr.Plot(
|
| 433 |
+
value=get_combined_leaderboard_with_radar(rank_data, {
|
| 434 |
+
"Super Mario Bros": True,
|
| 435 |
+
"Sokoban": True,
|
| 436 |
+
"2048": True,
|
| 437 |
+
"Candy Crash": True,
|
| 438 |
+
"Tetris (complete)": True,
|
| 439 |
+
"Tetris (planning only)": True
|
| 440 |
+
})[1],
|
| 441 |
+
label="Performance Visualization",
|
| 442 |
+
elem_classes="visualization-container"
|
| 443 |
+
)
|
| 444 |
+
|
| 445 |
+
# Game selection section
|
| 446 |
+
with gr.Row():
|
| 447 |
+
gr.Markdown("### 🎮 Game Selection")
|
| 448 |
+
with gr.Row():
|
| 449 |
+
# For each game, we have two checkboxes: one for overall and one for detailed view.
|
| 450 |
+
with gr.Column():
|
| 451 |
+
gr.Markdown("**🎮 Super Mario Bros**")
|
| 452 |
+
mario_overall = gr.Checkbox(label="Super Mario Bros Score", value=True)
|
| 453 |
+
mario_details = gr.Checkbox(label="Super Mario Bros Details", value=False)
|
| 454 |
+
with gr.Column():
|
| 455 |
+
gr.Markdown("**📦 Sokoban**")
|
| 456 |
+
sokoban_overall = gr.Checkbox(label="Sokoban Score", value=True)
|
| 457 |
+
sokoban_details = gr.Checkbox(label="Sokoban Details", value=False)
|
| 458 |
+
with gr.Column():
|
| 459 |
+
gr.Markdown("**🔢 2048**")
|
| 460 |
+
_2048_overall = gr.Checkbox(label="2048 Score", value=True)
|
| 461 |
+
_2048_details = gr.Checkbox(label="2048 Details", value=False)
|
| 462 |
+
with gr.Column():
|
| 463 |
+
gr.Markdown("**🍬 Candy Crash**")
|
| 464 |
+
candy_overall = gr.Checkbox(label="Candy Crash Score", value=True)
|
| 465 |
+
candy_details = gr.Checkbox(label="Candy Crash Details", value=False)
|
| 466 |
+
with gr.Column():
|
| 467 |
+
gr.Markdown("**🎯 Tetris (complete)**")
|
| 468 |
+
tetris_overall = gr.Checkbox(label="Tetris (complete) Score", value=True)
|
| 469 |
+
tetris_details = gr.Checkbox(label="Tetris (complete) Details", value=False)
|
| 470 |
+
with gr.Column():
|
| 471 |
+
gr.Markdown("**📋 Tetris (planning)**")
|
| 472 |
+
tetris_plan_overall = gr.Checkbox(label="Tetris (planning) Score", value=True)
|
| 473 |
+
tetris_plan_details = gr.Checkbox(label="Tetris (planning) Details", value=False)
|
| 474 |
+
|
| 475 |
+
# Time progression display and control buttons - Moved below game selection
|
| 476 |
+
with gr.Row():
|
| 477 |
+
with gr.Column(scale=2):
|
| 478 |
+
gr.Markdown("**⏰ Time Tracker**")
|
| 479 |
+
time_slider = gr.Slider(
|
| 480 |
+
minimum=0,
|
| 481 |
+
maximum=1,
|
| 482 |
+
value=1,
|
| 483 |
+
step=1,
|
| 484 |
+
label="Model Time Point",
|
| 485 |
+
info="Current Time: 03/25/2025"
|
| 486 |
+
)
|
| 487 |
+
with gr.Column(scale=1):
|
| 488 |
+
gr.Markdown("**🔄 Controls**")
|
| 489 |
+
clear_btn = gr.Button("Reset Filters", variant="secondary")
|
| 490 |
+
|
| 491 |
+
# Leaderboard table section
|
| 492 |
+
with gr.Row():
|
| 493 |
+
gr.Markdown("### 📋 Detailed Results")
|
| 494 |
+
with gr.Row():
|
| 495 |
+
leaderboard_board = gr.DataFrame(
|
| 496 |
+
value=get_combined_leaderboard(rank_data, {
|
| 497 |
+
"Super Mario Bros": True,
|
| 498 |
+
"Sokoban": True,
|
| 499 |
+
"2048": True,
|
| 500 |
+
"Candy Crash": True,
|
| 501 |
+
"Tetris (complete)": True,
|
| 502 |
+
"Tetris (planning only)": True
|
| 503 |
+
}),
|
| 504 |
+
interactive=True,
|
| 505 |
+
wrap=True,
|
| 506 |
+
label="Leaderboard"
|
| 507 |
+
)
|
| 508 |
+
|
| 509 |
+
# List of all checkboxes (in order)
|
| 510 |
+
checkbox_list = [mario_overall, mario_details,
|
| 511 |
+
sokoban_overall, sokoban_details,
|
| 512 |
+
_2048_overall, _2048_details,
|
| 513 |
+
candy_overall, candy_details,
|
| 514 |
+
tetris_overall, tetris_details,
|
| 515 |
+
tetris_plan_overall, tetris_plan_details]
|
| 516 |
+
|
| 517 |
+
# Initialize the leaderboard state when the app starts
|
| 518 |
+
clear_filters()
|
| 519 |
+
|
| 520 |
+
# Update both the leaderboard and visualization when checkboxes change
|
| 521 |
+
for checkbox in checkbox_list:
|
| 522 |
+
checkbox.change(
|
| 523 |
+
fn=update_leaderboard,
|
| 524 |
+
inputs=checkbox_list,
|
| 525 |
+
outputs=[leaderboard_board, visualization, gallery] + checkbox_list
|
| 526 |
+
)
|
| 527 |
+
|
| 528 |
+
# Update both when clear button is clicked
|
| 529 |
+
clear_btn.click(
|
| 530 |
+
fn=clear_filters,
|
| 531 |
+
inputs=[],
|
| 532 |
+
outputs=[leaderboard_board, visualization, gallery] + checkbox_list
|
| 533 |
+
)
|
| 534 |
+
|
| 535 |
+
return demo
|
| 536 |
+
|
| 537 |
+
if __name__ == "__main__":
|
| 538 |
+
demo_app = build_app()
|
| 539 |
+
# Add file serving configuration
|
| 540 |
+
demo_app.launch(debug=True, show_error=True, share=True)
|
leaderboard_utils.py
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import pandas as pd
|
| 2 |
+
import json
|
| 3 |
+
import numpy as np
|
| 4 |
+
|
| 5 |
+
# Define game order
|
| 6 |
+
GAME_ORDER = [
|
| 7 |
+
"Super Mario Bros",
|
| 8 |
+
"Sokoban",
|
| 9 |
+
"2048",
|
| 10 |
+
"Candy Crash",
|
| 11 |
+
"Tetris (complete)",
|
| 12 |
+
"Tetris (planning only)"
|
| 13 |
+
]
|
| 14 |
+
|
| 15 |
+
def get_organization(model_name):
|
| 16 |
+
m = model_name.lower()
|
| 17 |
+
if "claude" in m:
|
| 18 |
+
return "anthropic"
|
| 19 |
+
elif "gemini" in m:
|
| 20 |
+
return "google"
|
| 21 |
+
elif "o1" in m or "gpt" in m or "o3" in m:
|
| 22 |
+
return "openai"
|
| 23 |
+
elif "deepseek" in m:
|
| 24 |
+
return "deepseek"
|
| 25 |
+
else:
|
| 26 |
+
return "unknown"
|
| 27 |
+
|
| 28 |
+
def get_mario_leaderboard(rank_data):
|
| 29 |
+
data = rank_data.get("Super Mario Bros", {}).get("results", [])
|
| 30 |
+
df = pd.DataFrame(data)
|
| 31 |
+
df = df.rename(columns={
|
| 32 |
+
"model": "Player",
|
| 33 |
+
"progress": "Progress (current/total)",
|
| 34 |
+
"score": "Score",
|
| 35 |
+
"time_s": "Time (s)"
|
| 36 |
+
})
|
| 37 |
+
df["Organization"] = df["Player"].apply(get_organization)
|
| 38 |
+
df = df[["Player", "Organization", "Progress (current/total)", "Score", "Time (s)"]]
|
| 39 |
+
return df
|
| 40 |
+
|
| 41 |
+
def get_sokoban_leaderboard(rank_data):
|
| 42 |
+
data = rank_data.get("Sokoban", {}).get("results", [])
|
| 43 |
+
df = pd.DataFrame(data)
|
| 44 |
+
df = df.rename(columns={
|
| 45 |
+
"model": "Player",
|
| 46 |
+
"levels_cracked": "Levels Cracked",
|
| 47 |
+
"steps": "Steps"
|
| 48 |
+
})
|
| 49 |
+
df["Organization"] = df["Player"].apply(get_organization)
|
| 50 |
+
df = df[["Player", "Organization", "Levels Cracked", "Steps"]]
|
| 51 |
+
return df
|
| 52 |
+
|
| 53 |
+
def get_2048_leaderboard(rank_data):
|
| 54 |
+
data = rank_data.get("2048", {}).get("results", [])
|
| 55 |
+
df = pd.DataFrame(data)
|
| 56 |
+
df = df.rename(columns={
|
| 57 |
+
"model": "Player",
|
| 58 |
+
"score": "Score",
|
| 59 |
+
"steps": "Steps",
|
| 60 |
+
"time": "Time"
|
| 61 |
+
})
|
| 62 |
+
df["Organization"] = df["Player"].apply(get_organization)
|
| 63 |
+
df = df[["Player", "Organization", "Score", "Steps", "Time"]]
|
| 64 |
+
return df
|
| 65 |
+
|
| 66 |
+
def get_candy_leaderboard(rank_data):
|
| 67 |
+
data = rank_data.get("Candy Crash", {}).get("results", [])
|
| 68 |
+
df = pd.DataFrame(data)
|
| 69 |
+
df = df.rename(columns={
|
| 70 |
+
"model": "Player",
|
| 71 |
+
"score_runs": "Score Runs",
|
| 72 |
+
"average_score": "Average Score",
|
| 73 |
+
"steps": "Steps"
|
| 74 |
+
})
|
| 75 |
+
df["Organization"] = df["Player"].apply(get_organization)
|
| 76 |
+
df = df[["Player", "Organization", "Score Runs", "Average Score", "Steps"]]
|
| 77 |
+
return df
|
| 78 |
+
|
| 79 |
+
def get_tetris_leaderboard(rank_data):
|
| 80 |
+
data = rank_data.get("Tetris (complete)", {}).get("results", [])
|
| 81 |
+
df = pd.DataFrame(data)
|
| 82 |
+
df = df.rename(columns={
|
| 83 |
+
"model": "Player",
|
| 84 |
+
"score": "Score",
|
| 85 |
+
"steps_blocks": "Steps"
|
| 86 |
+
})
|
| 87 |
+
df["Organization"] = df["Player"].apply(get_organization)
|
| 88 |
+
df = df[["Player", "Organization", "Score", "Steps"]]
|
| 89 |
+
return df
|
| 90 |
+
|
| 91 |
+
def get_tetris_planning_leaderboard(rank_data):
|
| 92 |
+
data = rank_data.get("Tetris (planning only)", {}).get("results", [])
|
| 93 |
+
df = pd.DataFrame(data)
|
| 94 |
+
df = df.rename(columns={
|
| 95 |
+
"model": "Player",
|
| 96 |
+
"score": "Score",
|
| 97 |
+
"steps_blocks": "Steps"
|
| 98 |
+
})
|
| 99 |
+
df["Organization"] = df["Player"].apply(get_organization)
|
| 100 |
+
df = df[["Player", "Organization", "Score", "Steps"]]
|
| 101 |
+
return df
|
| 102 |
+
|
| 103 |
+
def calculate_rank_and_completeness(rank_data, selected_games):
|
| 104 |
+
# Dictionary to store DataFrames for each game
|
| 105 |
+
game_dfs = {}
|
| 106 |
+
|
| 107 |
+
# Get DataFrames for selected games
|
| 108 |
+
if selected_games.get("Super Mario Bros"):
|
| 109 |
+
game_dfs["Super Mario Bros"] = get_mario_leaderboard(rank_data)
|
| 110 |
+
if selected_games.get("Sokoban"):
|
| 111 |
+
game_dfs["Sokoban"] = get_sokoban_leaderboard(rank_data)
|
| 112 |
+
if selected_games.get("2048"):
|
| 113 |
+
game_dfs["2048"] = get_2048_leaderboard(rank_data)
|
| 114 |
+
if selected_games.get("Candy Crash"):
|
| 115 |
+
game_dfs["Candy Crash"] = get_candy_leaderboard(rank_data)
|
| 116 |
+
if selected_games.get("Tetris (complete)"):
|
| 117 |
+
game_dfs["Tetris (complete)"] = get_tetris_leaderboard(rank_data)
|
| 118 |
+
if selected_games.get("Tetris (planning only)"):
|
| 119 |
+
game_dfs["Tetris (planning only)"] = get_tetris_planning_leaderboard(rank_data)
|
| 120 |
+
|
| 121 |
+
# Get all unique players
|
| 122 |
+
all_players = set()
|
| 123 |
+
for df in game_dfs.values():
|
| 124 |
+
all_players.update(df["Player"].unique())
|
| 125 |
+
all_players = sorted(list(all_players))
|
| 126 |
+
|
| 127 |
+
# Create results DataFrame
|
| 128 |
+
results = []
|
| 129 |
+
for player in all_players:
|
| 130 |
+
player_data = {
|
| 131 |
+
"Player": player,
|
| 132 |
+
"Organization": get_organization(player)
|
| 133 |
+
}
|
| 134 |
+
ranks = []
|
| 135 |
+
games_played = 0
|
| 136 |
+
|
| 137 |
+
# Calculate rank and completeness for each game
|
| 138 |
+
for game in GAME_ORDER:
|
| 139 |
+
if game in game_dfs:
|
| 140 |
+
df = game_dfs[game]
|
| 141 |
+
if player in df["Player"].values:
|
| 142 |
+
games_played += 1
|
| 143 |
+
# Get player's score based on game type
|
| 144 |
+
if game == "Super Mario Bros":
|
| 145 |
+
player_score = df[df["Player"] == player]["Score"].iloc[0]
|
| 146 |
+
rank = len(df[df["Score"] > player_score]) + 1
|
| 147 |
+
elif game == "Sokoban":
|
| 148 |
+
# Parse Sokoban score string and get maximum level
|
| 149 |
+
levels_str = df[df["Player"] == player]["Levels Cracked"].iloc[0]
|
| 150 |
+
try:
|
| 151 |
+
# Split by semicolon, strip whitespace, filter empty strings, convert to integers
|
| 152 |
+
levels = [int(x.strip()) for x in levels_str.split(";") if x.strip()]
|
| 153 |
+
player_score = max(levels) if levels else 0
|
| 154 |
+
except:
|
| 155 |
+
player_score = 0
|
| 156 |
+
# Calculate rank based on maximum level
|
| 157 |
+
rank = len(df[df["Levels Cracked"].apply(
|
| 158 |
+
lambda x: max([int(y.strip()) for y in x.split(";") if y.strip()]) > player_score
|
| 159 |
+
)]) + 1
|
| 160 |
+
elif game == "2048":
|
| 161 |
+
player_score = df[df["Player"] == player]["Score"].iloc[0]
|
| 162 |
+
rank = len(df[df["Score"] > player_score]) + 1
|
| 163 |
+
elif game == "Candy Crash":
|
| 164 |
+
player_score = df[df["Player"] == player]["Average Score"].iloc[0]
|
| 165 |
+
rank = len(df[df["Average Score"] > player_score]) + 1
|
| 166 |
+
elif game == "Tetris (complete)":
|
| 167 |
+
player_score = df[df["Player"] == player]["Score"].iloc[0]
|
| 168 |
+
rank = len(df[df["Score"] > player_score]) + 1
|
| 169 |
+
elif game == "Tetris (planning only)":
|
| 170 |
+
player_score = df[df["Player"] == player]["Score"].iloc[0]
|
| 171 |
+
rank = len(df[df["Score"] > player_score]) + 1
|
| 172 |
+
|
| 173 |
+
ranks.append(rank)
|
| 174 |
+
player_data[f"{game} Score"] = player_score
|
| 175 |
+
else:
|
| 176 |
+
player_data[f"{game} Score"] = "_"
|
| 177 |
+
|
| 178 |
+
# Calculate average rank and completeness for sorting only
|
| 179 |
+
if ranks:
|
| 180 |
+
player_data["Sort Rank"] = round(np.mean(ranks), 2)
|
| 181 |
+
player_data["Games Played"] = games_played
|
| 182 |
+
else:
|
| 183 |
+
player_data["Sort Rank"] = float('inf')
|
| 184 |
+
player_data["Games Played"] = 0
|
| 185 |
+
|
| 186 |
+
results.append(player_data)
|
| 187 |
+
|
| 188 |
+
# Create DataFrame and sort by average rank and completeness
|
| 189 |
+
df_results = pd.DataFrame(results)
|
| 190 |
+
if not df_results.empty:
|
| 191 |
+
# Sort by average rank (ascending) and completeness (descending)
|
| 192 |
+
df_results = df_results.sort_values(
|
| 193 |
+
by=["Sort Rank", "Games Played"],
|
| 194 |
+
ascending=[True, False]
|
| 195 |
+
)
|
| 196 |
+
# Drop the sorting columns
|
| 197 |
+
df_results = df_results.drop(["Sort Rank", "Games Played"], axis=1)
|
| 198 |
+
|
| 199 |
+
return df_results
|
| 200 |
+
|
| 201 |
+
def get_combined_leaderboard(rank_data, selected_games):
|
| 202 |
+
return calculate_rank_and_completeness(rank_data, selected_games)
|
rank_data_03_25_2025.json
ADDED
|
@@ -0,0 +1,324 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"Super Mario Bros": {
|
| 3 |
+
"runs": 5,
|
| 4 |
+
"results": [
|
| 5 |
+
{
|
| 6 |
+
"model": "claude-3-7-sonnet-20250219",
|
| 7 |
+
"score": 710,
|
| 8 |
+
"progress": "1-1",
|
| 9 |
+
"time_s": 64.2,
|
| 10 |
+
"rank": 1
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"model": "gpt-4o-2024-11-20",
|
| 14 |
+
"score": 560,
|
| 15 |
+
"progress": "1-1",
|
| 16 |
+
"time_s": 58.6,
|
| 17 |
+
"rank": 2
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"model": "gemini-2.0-flash",
|
| 21 |
+
"score": 320,
|
| 22 |
+
"progress": "1-1",
|
| 23 |
+
"time_s": 51.8,
|
| 24 |
+
"rank": 3
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"model": "claude-3-5-haiku-20241022",
|
| 28 |
+
"score": 140,
|
| 29 |
+
"progress": "1-1",
|
| 30 |
+
"time_s": 76.4,
|
| 31 |
+
"rank": 4
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"model": "gpt-4.5-preview-2025-02-27",
|
| 35 |
+
"score": 160,
|
| 36 |
+
"progress": "1-1",
|
| 37 |
+
"time_s": 62.8,
|
| 38 |
+
"rank": 5
|
| 39 |
+
}
|
| 40 |
+
]
|
| 41 |
+
},
|
| 42 |
+
"2048": {
|
| 43 |
+
"runs": 1,
|
| 44 |
+
"results": [
|
| 45 |
+
{
|
| 46 |
+
"model": "claude-3-7-sonnet-20250219(thinking)",
|
| 47 |
+
"score": 256,
|
| 48 |
+
"steps": 114,
|
| 49 |
+
"time": ">200",
|
| 50 |
+
"rank": 1
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"model": "o1-2024-12-17",
|
| 54 |
+
"score": 256,
|
| 55 |
+
"steps": 116,
|
| 56 |
+
"time": ">200",
|
| 57 |
+
"rank": 2
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"model": "claude-3-7-sonnet-20250219",
|
| 61 |
+
"score": 256,
|
| 62 |
+
"steps": 130,
|
| 63 |
+
"time": "20:36",
|
| 64 |
+
"rank": 3
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"model": "deepseek-v3",
|
| 68 |
+
"score": 256,
|
| 69 |
+
"steps": 216,
|
| 70 |
+
"time": "54.02",
|
| 71 |
+
"rank": 4
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"model": "gemini-2.0-flash",
|
| 75 |
+
"score": 128,
|
| 76 |
+
"steps": 111,
|
| 77 |
+
"time": "18:43",
|
| 78 |
+
"rank": 5
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"model": "gemini-2.0-flash-thinking-exp-1219",
|
| 82 |
+
"score": 128,
|
| 83 |
+
"steps": 132,
|
| 84 |
+
"time": ">100",
|
| 85 |
+
"rank": 6
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"model": "gemini-2.5-pro-exp-03-25",
|
| 89 |
+
"score": 128,
|
| 90 |
+
"steps": 138,
|
| 91 |
+
"time": "169",
|
| 92 |
+
"rank": 7
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"model": "claude-3-5-sonnet-20241022",
|
| 96 |
+
"score": 64,
|
| 97 |
+
"steps": 92,
|
| 98 |
+
"time": "9:2",
|
| 99 |
+
"rank": 9
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"model": "gpt-4.5-preview-2025-02-27",
|
| 103 |
+
"score": 34,
|
| 104 |
+
"steps": 34,
|
| 105 |
+
"time": "8:25",
|
| 106 |
+
"rank": 10
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"model": "gpt-4o-2024-11-20",
|
| 110 |
+
"score": 16,
|
| 111 |
+
"steps": 21,
|
| 112 |
+
"time": "1:17",
|
| 113 |
+
"rank": 11
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"model": "Llama-4-Maverick-17B-128E-Instruct-FP8",
|
| 117 |
+
"score": 128,
|
| 118 |
+
"steps": 145,
|
| 119 |
+
"time": ">100",
|
| 120 |
+
"rank": 8
|
| 121 |
+
}
|
| 122 |
+
]
|
| 123 |
+
},
|
| 124 |
+
"Tetris (complete)": {
|
| 125 |
+
"runs": 3,
|
| 126 |
+
"results": [
|
| 127 |
+
{
|
| 128 |
+
"model": "claude-3-7-sonnet-20250219",
|
| 129 |
+
"score": 95,
|
| 130 |
+
"steps_blocks": 27,
|
| 131 |
+
"rank": 1
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"model": "claude-3-5-haiku-20241022",
|
| 135 |
+
"score": 90,
|
| 136 |
+
"steps_blocks": 25,
|
| 137 |
+
"rank": 2
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"model": "gemini-2.0-flash",
|
| 141 |
+
"score": 82,
|
| 142 |
+
"steps_blocks": 23,
|
| 143 |
+
"rank": 3
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"model": "gpt-4o-2024-11-20",
|
| 147 |
+
"score": 54,
|
| 148 |
+
"steps_blocks": 19,
|
| 149 |
+
"rank": 4
|
| 150 |
+
}
|
| 151 |
+
]
|
| 152 |
+
},
|
| 153 |
+
"Tetris (planning only)": {
|
| 154 |
+
"runs": 3,
|
| 155 |
+
"results": [
|
| 156 |
+
{
|
| 157 |
+
"model": "claude-3-7-sonnet-20250219",
|
| 158 |
+
"score": 110,
|
| 159 |
+
"steps_blocks": 29,
|
| 160 |
+
"rank": 1
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"model": "claude-3-5-haiku-20241022",
|
| 164 |
+
"score": 92,
|
| 165 |
+
"steps_blocks": 25,
|
| 166 |
+
"rank": 2
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"model": "gemini-2.0-flash",
|
| 170 |
+
"score": 87,
|
| 171 |
+
"steps_blocks": 24,
|
| 172 |
+
"rank": 3
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"model": "gpt-4o-2024-11-20",
|
| 176 |
+
"score": 56,
|
| 177 |
+
"steps_blocks": 20,
|
| 178 |
+
"rank": 4
|
| 179 |
+
}
|
| 180 |
+
]
|
| 181 |
+
},
|
| 182 |
+
"Candy Crash": {
|
| 183 |
+
"runs": 3,
|
| 184 |
+
"results": [
|
| 185 |
+
{
|
| 186 |
+
"model": "o3-mini-2025-01-31(medium)",
|
| 187 |
+
"score_runs": "90;109;120",
|
| 188 |
+
"average_score": 106.33,
|
| 189 |
+
"steps": 25,
|
| 190 |
+
"rank": 1
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"model": "o1-2024-12-17",
|
| 194 |
+
"score_runs": "96;114;83",
|
| 195 |
+
"average_score": 97.67,
|
| 196 |
+
"steps": 25,
|
| 197 |
+
"rank": 2
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"model": "deepseek-r1",
|
| 201 |
+
"score_runs": "62;108;105",
|
| 202 |
+
"average_score": 91.67,
|
| 203 |
+
"steps": 25,
|
| 204 |
+
"rank": 3
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"model": "gemini-2.5-pro-exp-03-25",
|
| 208 |
+
"score_runs": "50;36;68",
|
| 209 |
+
"average_score": 51.33,
|
| 210 |
+
"steps": 25,
|
| 211 |
+
"rank": 4
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"model": "claude-3-7-sonnet-20250219(thinking)",
|
| 215 |
+
"score_runs": "36;46;24",
|
| 216 |
+
"average_score": 35.33,
|
| 217 |
+
"steps": 25,
|
| 218 |
+
"rank": 5
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"model": "gemini-2.0-flash-thinking-exp-1219",
|
| 222 |
+
"score_runs": "0;15;39",
|
| 223 |
+
"average_score": 18,
|
| 224 |
+
"steps": 25,
|
| 225 |
+
"rank": 6
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"model": "claude-3-5-sonnet-20241022",
|
| 229 |
+
"score_runs": "3;0;0",
|
| 230 |
+
"average_score": 1,
|
| 231 |
+
"steps": 25,
|
| 232 |
+
"rank": 7
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"model": "deepseek-v3",
|
| 236 |
+
"score_runs": "0;0;0",
|
| 237 |
+
"average_score": 0,
|
| 238 |
+
"steps": 25,
|
| 239 |
+
"rank":9
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"model": "Llama-4-Maverick-17B-128E-Instruct-FP8",
|
| 243 |
+
"score_runs": "6;0;0",
|
| 244 |
+
"average_score": 2,
|
| 245 |
+
"steps": 25,
|
| 246 |
+
"rank": 8
|
| 247 |
+
}
|
| 248 |
+
]
|
| 249 |
+
},
|
| 250 |
+
"Sokoban": {
|
| 251 |
+
"runs": 3,
|
| 252 |
+
"results": [
|
| 253 |
+
{
|
| 254 |
+
"model": "o3-mini-2025-01-31(medium)",
|
| 255 |
+
"levels_cracked": "2; 3; 2",
|
| 256 |
+
"steps": "[17,52,68];[24,58,78,91];[19,44,64]",
|
| 257 |
+
"rank": 1
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"model": "gemini-2.5-pro-exp-03-25",
|
| 261 |
+
"levels_cracked": "2;2;3",
|
| 262 |
+
"steps": "[23, 46, 79]; [20,50,77]; [26,95,125,175]",
|
| 263 |
+
"rank": 2
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"model": "claude-3-7-sonnet-20250219(thinking)",
|
| 267 |
+
"levels_cracked": "1; 2; 0",
|
| 268 |
+
"steps": "[17,35];[15,40,43];[4]",
|
| 269 |
+
"rank": 3
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"model": "o1-2024-12-17",
|
| 273 |
+
"levels_cracked": "1; 1; 1",
|
| 274 |
+
"steps": null,
|
| 275 |
+
"rank": 4
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"model": "deepseek-r1",
|
| 279 |
+
"levels_cracked": "1; 0; 1",
|
| 280 |
+
"steps": "[19,42];[13];[19,36]",
|
| 281 |
+
"note": "stuck",
|
| 282 |
+
"rank": 5
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"model": "o1-mini-2024-09-12",
|
| 286 |
+
"levels_cracked": "0;1;0",
|
| 287 |
+
"steps": null,
|
| 288 |
+
"rank": 6
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"model": "gemini-2.0-flash-thinking-exp-1219",
|
| 292 |
+
"levels_cracked": "0; 0; 0",
|
| 293 |
+
"steps": "[23]; [14]; [14]",
|
| 294 |
+
"rank": 7
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"model": "gpt-4o-2024-11-20",
|
| 298 |
+
"levels_cracked": "0; 0; 0",
|
| 299 |
+
"steps": "[68];[105];[168]",
|
| 300 |
+
"note": "stuck in a loop",
|
| 301 |
+
"rank": 8
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"model": "claude-3-5-sonnet-20241022",
|
| 305 |
+
"levels_cracked": "0; 0; 0",
|
| 306 |
+
"steps": "[21]; [30]; [51]",
|
| 307 |
+
"note": "stuck in a loop",
|
| 308 |
+
"rank": 9
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"model": "deepseek-v3",
|
| 312 |
+
"levels_cracked": "0; 0; 0",
|
| 313 |
+
"steps": "[9]; [47]; [64]",
|
| 314 |
+
"rank": 10
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"model": "Llama-4-Maverick-17B-128E-Instruct-FP8",
|
| 318 |
+
"levels_cracked": "0;0;0",
|
| 319 |
+
"steps": "[5]",
|
| 320 |
+
"rank": 11
|
| 321 |
+
}
|
| 322 |
+
]
|
| 323 |
+
}
|
| 324 |
+
}
|