nostromo

app.py

@@ -1,18 +1,12 @@
-"""
-Gradio requires input to be fed in a very peculiar way and does not provide too much flexibility - don't expect from this demo too much. The backbone had to be adjusted to work on hugging face spaces. Go see https://github.com/PiotrAntoniak/QuestionAnswering for a prettier version utilizing streamlit. 
-"""
-
-
 import gradio as gr
  
 description = """Do you have a long document and a bunch of questions that can be answered given the data in this file? 
 Fear not for this demo is for you. 
 Upload your pdf, ask your questions and wait for the magic to happen.
 DISCLAIMER: I do no have idea what happens to the pdfs that you upload and who has access to them so make sure there is nothing confidential there. 
-""" 
+"""
 title = "QA answering from a pdf."
 
-from datetime import datetime
 import numpy as np
 import time
 import hashlib
@@ -57,7 +51,7 @@ def encode_docs(docs,maxlen = 64, stride = 32):
     spans = []
     file_names = []
     name, text = docs
-    temp_text = ""
+    
     text = text.split(" ")
     if len(text) < maxlen:
         text = " ".join(text)
@@ -92,7 +86,6 @@ def encode_docs(docs,maxlen = 64, stride = 32):
     return embeddings, spans, file_names
    
 def predict(query,data):
-    print(datetime.today().strftime('%Y-%m-%d %H:%M:%S'))
     name_to_save = data.name.split("/")[-1].split(".")[0][:-8]
     k=20
     st = str([query,name_to_save])
@@ -175,7 +168,6 @@ def predict(query,data):
     print(df)
     print("time: "+ str(time.time()-start))
     
-    
     with open("HISTORY.txt","a", encoding = "utf-8") as f:
         f.write(hist)
         f.write(" " + str(current_time))
@@ -195,20 +187,21 @@ def predict(query,data):
     
     return list_outputs
 
-
-
 iface = gr.Interface(examples = [
         ["How high is the highest mountain?","China.pdf"], 
-        ["Where is the highest  mountain?","China.pdf"]
+        ["Where does UK prime minister live?","London.pdf"]
     ],
     
     fn =predict,
-    inputs = [gr.Textbox(),
-              gr.File(),
+    inputs = [gr.inputs.Textbox(default="What is Open-domain question answering?"),
+              gr.inputs.File(),
     ],
-    outputs = 'text',
+    outputs = [
+        gr.outputs.Carousel(['text']),
+            ],
     description=description,
-    title = title
-                     )
+    title = title,
+allow_flagging ="manual",flagging_options = ["correct","wrong"],
+                     allow_screenshot=False)
 
-iface.launch()
+iface.launch(share = True,enable_queue=True, show_error =True)

requirements.txt

@@ -3,5 +3,4 @@ textract
 scipy
 pandas
 numpy
-transformers
-gradio==3.0.20
+transformers