Update app.py

app.py

@@ -45,6 +45,8 @@ set_state_if_absent("summaries", None)
 document_store, summarizer, preprocessor = start_haystack()
 
 st.title('TL;DR with Haystack')
+image = Image.open('header-image.png')
+st.image(image)
 
 st.markdown( """
 This Summarization demo uses a [Haystack TransformerSummarizer node](https://haystack.deepset.ai/pipeline_nodes/summarizer). You can upload a PDF file, which will be converted to text with the [Haystack PDFtoTextConverter](https://haystack.deepset.ai/reference/file-converters#pdftotextconverter). In this demo, we produce 1 summary for the whole file you upload. So, the TransformerSummarizer treats the whole thing as one string, which means along with the model limitations, PDFs that have a lot of unneeded text at the beginning produce poor results. For best results, upload a document that has minimal intro and tables at the top.