edited UI

.gitignore

@@ -0,0 +1 @@
+__pycache__

.streamlit/config.toml

@@ -1,6 +1,6 @@
 [theme]
 primaryColor = "#E694FF"
 backgroundColor = "#DCDCDC"
-secondaryBackgroundColor = "#0083B8"
+secondaryBackgroundColor = "#B8E3F5"
 textColor = "#00172B"
 font = "sans-serif"

app.py

@@ -18,12 +18,11 @@ st.markdown(
     """
 
 # Koya Recommeder System
-### 👋 Welcome to the to the Koya recommendation system. This system recommeds an LLM for you when you provide a sample sentence in your target language and select a list of models. 
+#### 👋 Welcome to the to the Koya recommendation system. This system recommeds an LLM for you when you provide a sample sentence in your target language and select a list of models. 
 You can try it below \n\n\n"""
 )
 
 
-
 @st.cache
 def get_model_infos(multilingual="multilingual"):
     api = HfApi()
@@ -209,15 +208,40 @@ def set_seed():
     np.random.seed(2023)
     torch.manual_seed(2023)
 
+
 with st.sidebar:
-    st.selectbox("df", [1,2,3,4])
+
+    st.image("Koya_Presentation-removebg-preview.png")
+    st.subheader("Abstract")
+    st.markdown(
+        """
+        <div style="text-align: justify">
+        <h6> Pretrained large language models (LLMs) are widely used for various downstream tasks in different languages. However, selecting the best
+         LLM (from a large set of potential LLMs) for a given downstream task and language is a challenging and computationally expensive task, making 
+         the efficient use of LLMs difficult for low-compute communities. To address this challenge, we present Koya, a recommender system built to assist 
+         researchers and practitioners in choosing the right LLM for their task and language, without ever having to finetune the LLMs. Koya is built with 
+         the Koya Pseudo-Perplexity (KPPPL), our adaptation of the pseudo perplexity, and ranks LLMs in order of compatibility with the language of interest, 
+         making it easier and cheaper to choose the most compatible LLM. By evaluating Koya using five pretrained LLMs and three African languages 
+         (Yoruba, Kinyarwanda, and Amharic), we show an average recommender accuracy of 95%, demonstrating its effectiveness. Koya aims to offer 
+         an easy to use (through a simple web interface accessible at https://huggingface.co/spaces/koya-recommender/system), cost-effective, fast and 
+         efficient tool to assist researchers and practitioners with low or limited compute access.</h6>
+        </div>
+
+        """,
+        unsafe_allow_html=True,
+    )
+    url = "https://share.streamlit.io/mesmith027/streamlit_webapps/main/MC_pi/streamlit_app.py"
+    st.write("check out the paper [here](%s)" % url)
+
 footer()
 
 sentence = st.text_input("Please input a sample sentence in the target language")
 
 models = get_model_infos(multilingual=None)
 selected_models = st.multiselect(
-    "Select of number of models you would like to compare", models["id"]
+    "Select of number of models you would like to compare",
+    models["id"],
+    max_selections=3,
 )
 
 run = st.button("Get Scores")

requirements.txt

@@ -1,5 +1,5 @@
 pandas
-streamlit
+streamlit==1.12.0
 numpy
 torch
 huggingface_hub

styling.py

@@ -68,7 +68,7 @@ def layout(*args):
 def footer():
     myargs = [
         " with ❤️ by ",
-        link("https://twitter.com/AbrahamOwos", "@AbrahamOwos "), "and ", link("https://twitter.com/AbrahamOwos", "@Chrisjay"),
+        link("https://twitter.com/AbrahamOwos", "@AbrahamOwos "), "and ", link("https://twitter.com/ChrisEmezue", "@ChrisEmezue"),
         br(),
         " ⚙️ Have a suggestion? Hit us up" ,
     ]