Stefano Fiorucci commited on
Commit
be93ed6
1 Parent(s): c557ead

random questions and other...

Browse files
app.py CHANGED
@@ -38,7 +38,13 @@ def start_haystack():
38
  )
39
  reader = FARMReader(model_name_or_path="deepset/roberta-base-squad2", use_gpu=True)
40
  pipe = ExtractiveQAPipeline(reader, retriever)
41
- return pipe
 
 
 
 
 
 
42
 
43
  def set_state_if_absent(key, value):
44
  if key not in st.session_state:
@@ -48,21 +54,13 @@ def query(pipe, question):
48
  """Run query and get answers"""
49
  return (pipe.run(question, params={"Retriever": {"top_k": 10}, "Reader": {"top_k": 5}}), None)
50
 
51
- def get_backlink(result) -> Tuple[Optional[str], Optional[str]]:
52
- if result.get("document", None):
53
- doc = result["document"]
54
- if isinstance(doc, dict):
55
- if doc.get("meta", None):
56
- if isinstance(doc["meta"], dict):
57
- if doc["meta"].get("url", None) and doc["meta"].get("name", None):
58
- return doc["meta"]["url"], doc["meta"]["name"]
59
- return None, None
60
 
61
  def main():
62
  # st.set_page_config(page_title='Who killed Laura Palmer?',
63
  # page_icon="https://static.wikia.nocookie.net/twinpeaks/images/4/4a/Site-favicon.ico/revision/latest?cb=20210710003705")
64
 
65
  pipe=start_haystack()
 
66
 
67
  # Persistent state
68
  set_state_if_absent('question', "Where is Twin Peaks?")
@@ -157,15 +155,14 @@ and see if the AI ​​can find an answer...
157
  # Run button
158
  run_pressed = col1.button("Run")
159
 
160
- df=''
161
  # Get next random question from the CSV
162
  if col2.button("Random question"):
163
  reset_results()
164
- new_row = df.sample(1)
165
- while new_row["Question Text"].values[0] == st.session_state.question: # Avoid picking the same question twice (the change is not visible on the UI)
166
- new_row = df.sample(1)
167
- st.session_state.question = new_row["Question Text"].values[0]
168
- st.session_state.answer = new_row["Answer"].values[0]
169
  st.session_state.random_question_requested = True
170
  # Re-runs the script setting the random question as the textbox value
171
  # Unfortunately necessary as the Random Question button is _below_ the textbox
@@ -220,12 +217,10 @@ and see if the AI ​​can find an answer...
220
  # Hack due to this bug: https://github.com/streamlit/streamlit/issues/3190
221
  st.write(markdown("- ..."+context[:start_idx] + str(annotation(answer, "ANSWER", "#3e1c21")) + context[end_idx:]+"..."), unsafe_allow_html=True)
222
  source = ""
223
- url, title = get_backlink(result)
224
- if url and title:
225
- source = f"[{result['document']['meta']['title']}]({result['document']['meta']['url']})"
226
- else:
227
- source = f"{result['source']}"
228
- st.markdown(f"**Score:** {result['relevance']} - **Source:** {source}")
229
  else:
230
  st.info("🤔    Haystack is unsure whether any of the documents contain an answer to your question. Try to reformulate it!")
231
  main()
38
  )
39
  reader = FARMReader(model_name_or_path="deepset/roberta-base-squad2", use_gpu=True)
40
  pipe = ExtractiveQAPipeline(reader, retriever)
41
+ return pipe
42
+
43
+ @st.cache()
44
+ def load_questions():
45
+ with open('./data/questions.txt') as fin:
46
+ questions = [line.strip() for line in fin.readlines()]
47
+ return questions
48
 
49
  def set_state_if_absent(key, value):
50
  if key not in st.session_state:
54
  """Run query and get answers"""
55
  return (pipe.run(question, params={"Retriever": {"top_k": 10}, "Reader": {"top_k": 5}}), None)
56
 
 
 
 
 
 
 
 
 
 
57
 
58
  def main():
59
  # st.set_page_config(page_title='Who killed Laura Palmer?',
60
  # page_icon="https://static.wikia.nocookie.net/twinpeaks/images/4/4a/Site-favicon.ico/revision/latest?cb=20210710003705")
61
 
62
  pipe=start_haystack()
63
+ questions = load_questions()
64
 
65
  # Persistent state
66
  set_state_if_absent('question', "Where is Twin Peaks?")
155
  # Run button
156
  run_pressed = col1.button("Run")
157
 
 
158
  # Get next random question from the CSV
159
  if col2.button("Random question"):
160
  reset_results()
161
+ question = random.choice(questions)
162
+ while question == st.session_state.question: # Avoid picking the same question twice (the change is not visible on the UI)
163
+ question = random.choice(questions)
164
+ st.session_state.question = question
165
+ # st.session_state.answer = new_row["Answer"].values[0]
166
  st.session_state.random_question_requested = True
167
  # Re-runs the script setting the random question as the textbox value
168
  # Unfortunately necessary as the Random Question button is _below_ the textbox
217
  # Hack due to this bug: https://github.com/streamlit/streamlit/issues/3190
218
  st.write(markdown("- ..."+context[:start_idx] + str(annotation(answer, "ANSWER", "#3e1c21")) + context[end_idx:]+"..."), unsafe_allow_html=True)
219
  source = ""
220
+ name = result['meta']['name']
221
+ url = result['meta']['url']
222
+ source = f"[{name}]({url})"
223
+ st.markdown(f"**Score:** {result['score']:.2f} - **Source:** {source}")
 
 
224
  else:
225
  st.info("🤔    Haystack is unsure whether any of the documents contain an answer to your question. Try to reformulate it!")
226
  main()
data/questions.txt ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Who killed Laura Palmer?
2
+ Who found Laura's body?
3
+ What is garmonbozia?
4
+ What is a doppelganger?
5
+ Who is Teresa Banks?
6
+ What was found under Laura's nail?
7
+ Who is Dale Cooper's fiancèe?
8
+ Who is Mrs Tremond
9
+ Who is Ray Monroe?
10
+ Where is Twin Peaks
11
+ Who is the real owner of One Eyed Jack?
12
+ Who works as a croupier?
13
+ Why was Laura Palmer killed?
14
+ Who is Evelyn Marsh?
15
+ Who is the log lady?
16
+ Who is Bobby Briggs' father?
17
+ who is Susan Hurley
18
+ Who is Mike
{index_creation → notebooks}/index_creation.ipynb RENAMED
File without changes