Update app.py

app.py

@@ -17,12 +17,18 @@ with st.echo():
     eng_num_tokens = dataset.map(lambda x: {'num_tokens':len(tokenizer(x['sentence_eng_Latn'])['input_ids'])})['num_tokens']
     ukr_num_tokens = dataset.map(lambda x: {'num_tokens':len(tokenizer(x['sentence_ukr_Cyrl'])['input_ids'])})['num_tokens']
 
-fig, (axl, axr) = plt.subplots(1, 2, figsize=(10,3))
-axl.hist(eng_num_tokens)
-axl.set_title(f'eng mistral tokens ({np.sum(eng_num_tokens)} total)')
-axr.hist(ukr_num_tokens)
-axr.set_title(f'ukr mistral tokens ({np.sum(ukr_num_tokens)} total)')
-st.pyplot(fig)
+with st.sidebar:
+    fig, (axl, axr) = plt.subplots(2, 1, figsize=(3,10))
+    axl.hist(eng_num_tokens)
+    axl.set_title(f'eng mistral tokens ({np.sum(eng_num_tokens)} total)')
+    axr.hist(ukr_num_tokens)
+    axr.set_title(f'ukr mistral tokens ({np.sum(ukr_num_tokens)} total)')
+    st.pyplot(fig)
+
+keyword = st.text_input("Filter by text", value="")
 
-st.dataframe(pd.DataFrame(dataset))
+if not keyword:
+    st.dataframe(pd.DataFrame(dataset))
+else:
+    st.dataframe(pd.DataFrame(dataset.filter(lambda x: keyword in x['sentence_eng_Latn'] or keyword in x['sentence_ukr_Cyrl'])))