Spaces:
Paused
Paused
Synced repo using 'sync_with_huggingface' Github Action
Browse files
app.py
CHANGED
|
@@ -503,11 +503,11 @@ def main():
|
|
| 503 |
|
| 504 |
with col1:
|
| 505 |
# Display header
|
| 506 |
-
st.title("
|
| 507 |
st.markdown(
|
| 508 |
"""
|
| 509 |
This app uses multiple LLMs (GPT-4o, Claude 3.7, and DeepSeekV3.0) to answer your questions.
|
| 510 |
-
The world's best LLM-as-a-Judge, [Selene](https://www.atla-ai.com/api), evaluates each response on accuracy, relevance, clarity, and
|
| 511 |
"""
|
| 512 |
)
|
| 513 |
|
|
|
|
| 503 |
|
| 504 |
with col1:
|
| 505 |
# Display header
|
| 506 |
+
st.title("🏛️ LLMs on Trial")
|
| 507 |
st.markdown(
|
| 508 |
"""
|
| 509 |
This app uses multiple LLMs (GPT-4o, Claude 3.7, and DeepSeekV3.0) to answer your questions.
|
| 510 |
+
The world's best LLM-as-a-Judge, [Selene](https://www.atla-ai.com/api), evaluates each response on accuracy, relevance, clarity, and helpfulness, and the best one is selected and refined if needed (< 4.0 score).
|
| 511 |
"""
|
| 512 |
)
|
| 513 |
|