Renming Zhang
commited on
Commit
β’
43feb65
1
Parent(s):
55aff67
added image
Browse files- app.py +11 -2
- assets/logo.png +0 -0
- src/display/about.py +0 -1
app.py
CHANGED
@@ -24,6 +24,7 @@ from src.display.utils import (
|
|
24 |
Precision
|
25 |
)
|
26 |
from src.envs import API, EVAL_REQUESTS_PATH, EVAL_RESULTS_PATH, H4_TOKEN, IS_PUBLIC, QUEUE_REPO, REPO_ID, RESULTS_REPO
|
|
|
27 |
# from src.populate import get_evaluation_queue_df, get_leaderboard_df
|
28 |
# from src.submission.submit import add_new_eval
|
29 |
# from src.tools.collections import update_collections
|
@@ -207,7 +208,7 @@ class LLM_Model:
|
|
207 |
|
208 |
|
209 |
games = ["Breakthrough", "Connect Four", "Blind Auction", "Kuhn Poker",
|
210 |
-
"Liar's Dice", "Negotiation", "Nim", "Pig", "Iterated
|
211 |
|
212 |
# models = ["gpt-35-turbo-1106", "gpt-4", "Llama-2-70b-chat-hf", "CodeLlama-34b-Instruct-hf",
|
213 |
# "CodeLlama-70b-Instruct-hf", "Mistral-7B-Instruct-v01", "Mistral-7B-OpenOrca"]
|
@@ -217,10 +218,18 @@ games = ["Breakthrough", "Connect Four", "Blind Auction", "Kuhn Poker",
|
|
217 |
|
218 |
demo = gr.Blocks(css=custom_css)
|
219 |
|
|
|
|
|
|
|
|
|
|
|
|
|
220 |
with demo:
|
|
|
|
|
221 |
gr.HTML(TITLE)
|
222 |
-
gr.Markdown(INTRODUCTION_TEXT, elem_classes="markdown-text")
|
223 |
|
|
|
224 |
with gr.Tabs(elem_classes="tab-buttons") as tabs:
|
225 |
with gr.TabItem("π
GTBench", elem_id="llm-benchmark-tab-table", id=0):
|
226 |
with gr.Row():
|
|
|
24 |
Precision
|
25 |
)
|
26 |
from src.envs import API, EVAL_REQUESTS_PATH, EVAL_RESULTS_PATH, H4_TOKEN, IS_PUBLIC, QUEUE_REPO, REPO_ID, RESULTS_REPO
|
27 |
+
from PIL import Image
|
28 |
# from src.populate import get_evaluation_queue_df, get_leaderboard_df
|
29 |
# from src.submission.submit import add_new_eval
|
30 |
# from src.tools.collections import update_collections
|
|
|
208 |
|
209 |
|
210 |
games = ["Breakthrough", "Connect Four", "Blind Auction", "Kuhn Poker",
|
211 |
+
"Liar's Dice", "Negotiation", "Nim", "Pig", "Iterated Prisoner's Dilemma", "Tic-Tac-Toe"]
|
212 |
|
213 |
# models = ["gpt-35-turbo-1106", "gpt-4", "Llama-2-70b-chat-hf", "CodeLlama-34b-Instruct-hf",
|
214 |
# "CodeLlama-70b-Instruct-hf", "Mistral-7B-Instruct-v01", "Mistral-7B-OpenOrca"]
|
|
|
218 |
|
219 |
demo = gr.Blocks(css=custom_css)
|
220 |
|
221 |
+
|
222 |
+
def load_image(image_path):
|
223 |
+
image = Image.open(image_path)
|
224 |
+
return image
|
225 |
+
|
226 |
+
|
227 |
with demo:
|
228 |
+
gr.Image("./assets/logo.png", height="200px", width="200px",
|
229 |
+
show_download_button=False, container=False)
|
230 |
gr.HTML(TITLE)
|
|
|
231 |
|
232 |
+
gr.Markdown(INTRODUCTION_TEXT, elem_classes="markdown-text")
|
233 |
with gr.Tabs(elem_classes="tab-buttons") as tabs:
|
234 |
with gr.TabItem("π
GTBench", elem_id="llm-benchmark-tab-table", id=0):
|
235 |
with gr.Row():
|
assets/logo.png
ADDED
src/display/about.py
CHANGED
@@ -1,7 +1,6 @@
|
|
1 |
# from src.display.utils import ModelType
|
2 |
|
3 |
TITLE = """
|
4 |
-
<embed src="../../assets/logo.pdf" width="200px" height="200px" />
|
5 |
<h1 align="center" id="space-title">GTBench: Uncovering the Strategic Reasoning Limitation of LLMs via Game-Theoretic Evaluations</h1>"""
|
6 |
|
7 |
INTRODUCTION_TEXT = """
|
|
|
1 |
# from src.display.utils import ModelType
|
2 |
|
3 |
TITLE = """
|
|
|
4 |
<h1 align="center" id="space-title">GTBench: Uncovering the Strategic Reasoning Limitation of LLMs via Game-Theoretic Evaluations</h1>"""
|
5 |
|
6 |
INTRODUCTION_TEXT = """
|