Added topk and topp

app.py

@@ -23,7 +23,9 @@ The ship was on early Wednesday ordered to return to the Kai Tak Cruise Terminal
 sample_texts = [[text_1 ], [text_2]]
 
 desc = """
-<p style='text-align: center; color: #FF7F50'>This is an abstractive text summarizer app using fine-tuned bart-large-cnn model. The abstractive approach involves rephrasing the complete document while capturing the complete meaning of the document. This type of summarization provides more human-like summary. Note: For faster summaries input smaller texts. Sample Text input is provided for you at the bottom!</p>
+<p style='text-align: center; color: #FF7F50'>This is an abstractive text summarizer app using fine-tuned bart-large-cnn model. The abstractive approach involves rephrasing the complete document while capturing the complete meaning of the document. This type of summarization provides more human-like summary. 
+<p style='text-align: center; color: #FF7F50'> Note: For faster summaries input smaller texts.</'p>
+<p style='text-align: center; color: #FF7F50'>Sample text inputs are provided at the bottom!</p>
 """
 
 
@@ -35,7 +37,7 @@ model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
 def summarize(inp):
     inp = inp.replace('\n','')
     inp = tokenizer.encode(inp, return_tensors='pt', max_length=1024)
-    summary_ids = model.generate(inp, num_beams=4, max_length=150, early_stopping=True)
+    summary_ids = model.generate(inp, num_beams=4, max_length=150, early_stopping=True, do_sample=True, top_k=50, top_p=0.95)
     summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)
     return summary