Upload 7 files

app.py

@@ -0,0 +1,45 @@
+import streamlit as st
+
+from joblib import load
+import pandas as pd
+import plotly.express as px
+
+
+reduced_embeddings = load("embeddings_tsne.joblib")
+combined_sentences = load("sentence_list.joblib")
+languages = load("language_list.joblib")
+clusters = load("clusters.joblib")
+
+def main():
+    st.title("LASER Multilingual Sentence Embeddings Visualization")
+
+    df = pd.DataFrame({
+        'TSNE Component 1': reduced_embeddings[:, 0],
+        'TSNE Component 2': reduced_embeddings[:, 1],
+        'Language': languages,
+        'Sentence': combined_sentences,
+        'Cluster': ['Cluster {}'.format(cluster) for cluster in clusters]
+    })
+
+    select_all = st.checkbox("Select All Clusters")
+
+    unique_clusters = df['Cluster'].unique()
+
+    # Active only if 'Select All Clusters' is unchecked
+    if not select_all:
+        selected_clusters = st.multiselect("Select clusters to display", unique_clusters, default=unique_clusters[:10])
+    else:
+        selected_clusters = unique_clusters
+
+    filtered_df = df[df['Cluster'].isin(selected_clusters)]
+
+    fig = px.scatter(filtered_df, x='TSNE Component 1', y='TSNE Component 2', 
+                     color='Language', hover_data=['Sentence', 'Cluster'])
+    fig.update_layout(title="Multilingual Sentence Embeddings Visualization",
+                      xaxis_title="TSNE Component 1", yaxis_title="TSNE Component 2",
+                      legend_title="Language")
+
+    st.plotly_chart(fig, use_container_width=True)
+
+if __name__ == "__main__":
+    main()

clusters.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfebc019aa4c59a92263fd8827382a7740cc3b834ee0b36718862eb9f0269363
+size 1025

embeddings_tsne.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2a4a720720a90789b2ab0243275d129d85106bc1183226927285b7748376933
+size 1825

language_list.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ecf1dc86b022f7223afee99db1ccd1aae834733bd6a13c0ef0962fa81ed96a52
+size 495

requirements.txt

@@ -0,0 +1,3 @@
+plotly
+streamlit
+joblib

sentence_list.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84d94a24b9bbde9827cf8a75517d09c3add35717e20cae1ced20f65d9aee3999
+size 7669