minor updates

app.py

@@ -42,8 +42,8 @@ if menu == "Contributions & Evaluation":
 | Model     | [FullEvaluation](https://docs.google.com/spreadsheets/d/1vXJrIg38cEaKjOG5y4I4PQwAQFUmCkohbViJ9zj_Emg/edit#gid=1809754143) Average     | 20Newsgroups Clustering | StackOverflow DupQuestions | Twitter SemEval2015    |
 |-----------|---------------------------------------|-------|-------|-------|
 | paraphrase-mpnet-base-v2 (previous SOTA)  | 67.97 | 47.79 | 49.03 | 72.36 |
-| **all_datasets_v3_roberta-large (400k steps)** | **70.22** | 50.12 | 52.18 | 75.28 |
-| **all_datasets_v3_mpnet-base (440k steps)** | **70.01**   | 50.22 | 52.24 | 76.27 |
+| **all_datasets_v3_roberta-large (400k steps)** | **70.22** | **50.12** | **52.18** | **75.28** |
+| **all_datasets_v3_mpnet-base (440k steps)** | **70.01**   | **50.22** | **52.24** | **76.27** |
 ''')
 elif menu == "Sentence Similarity":
     st.header('Sentence Similarity')
@@ -54,7 +54,7 @@ metric between our main sentence and the others.
 
 For more cool information on sentence embeddings, see the [sBert project](https://www.sbert.net/examples/applications/computing-embeddings/README.html).
 ''')
-    select_models = st.multiselect("Choose models", options=list(MODELS_ID), default=list(MODELS_ID)[0])
+    select_models = st.multiselect("Choose models", options=list(MODELS_ID), default=list(MODELS_ID))
 
     anchor = st.text_input(
         'Please enter here the main text you want to compare:',
@@ -93,8 +93,8 @@ elif menu == "Asymmetric QA":
 **Instructions**: You can compare the Answer likeliness of a given Query with answer candidates of your choice. In the 
 background, we'll create an embedding for each answer, and then we'll use the cosine similarity function to calculate a 
 similarity metric between our query sentence and the others.
-`mpnet_asymmetric_qa` model works best for hard-negative answers or distinguishing similar queries due to separate models 
-applied for encoding questions and answers.
+`mpnet_asymmetric_qa` model works best for hard-negative answers or distinguishing answers that are actually questions 
+due to separate models applied for encoding questions and answers.
 
 For more cool information on sentence embeddings, see the [sBert project](https://www.sbert.net/examples/applications/computing-embeddings/README.html).
 ''')