lfoppiano commited on
Commit
9c4d6ae
โ€ข
1 Parent(s): 0f074cc

reorganise a bit the interface

Browse files
Files changed (1) hide show
  1. streamlit_app.py +35 -18
streamlit_app.py CHANGED
@@ -51,11 +51,24 @@ if 'ner_processing' not in st.session_state:
51
  if 'uploaded' not in st.session_state:
52
  st.session_state['uploaded'] = False
53
 
 
 
 
 
 
 
 
 
 
 
 
 
54
  def new_file():
55
  st.session_state['loaded_embeddings'] = None
56
  st.session_state['doc_id'] = None
57
  st.session_state['uploaded'] = True
58
 
 
59
  # @st.cache_resource
60
  def init_qa(model):
61
  if model == 'chatgpt-3.5-turbo':
@@ -134,59 +147,63 @@ def play_old_messages():
134
  # is_api_key_provided = st.session_state['api_key']
135
 
136
  with st.sidebar:
137
- st.markdown(
138
- ":warning: Do not upload sensitive data. We **temporarily** store text from the uploaded PDF documents solely for the purpose of processing your request, and we **do not assume responsibility** for any subsequent use or handling of the data submitted to third parties LLMs.")
139
-
140
  st.session_state['model'] = model = st.radio(
141
  "Model",
142
  ("chatgpt-3.5-turbo", "mistral-7b-instruct-v0.1"), # , "llama-2-70b-chat"),
143
  index=1,
144
  captions=[
145
  "ChatGPT 3.5 Turbo + Ada-002-text (embeddings)",
146
- "Mistral-7B-Instruct-V0.1 + Sentence BERT (embeddings)"
147
- # "LLama2-70B-Chat + Sentence BERT (embeddings)",
148
  ],
149
  help="Select the LLM model and embeddings you want to use.",
150
  disabled=st.session_state['doc_id'] is not None or st.session_state['uploaded'])
151
 
 
 
 
152
  if model == 'mistral-7b-instruct-v0.1' or model == 'llama-2-70b-chat':
153
  if 'HUGGINGFACEHUB_API_TOKEN' not in os.environ:
154
  api_key = st.text_input('Huggingface API Key', type="password")
155
 
156
  st.markdown(
157
- "Get it for [Open AI](https://platform.openai.com/account/api-keys) or [Huggingface](https://huggingface.co/docs/hub/security-tokens)")
158
  else:
159
  api_key = os.environ['HUGGINGFACEHUB_API_TOKEN']
160
 
161
  if api_key:
162
  # st.session_state['api_key'] = is_api_key_provided = True
163
- with st.spinner("Preparing environment"):
164
- st.session_state['api_keys']['mistral-7b-instruct-v0.1'] = api_key
165
- if 'HUGGINGFACEHUB_API_TOKEN' not in os.environ:
166
- os.environ["HUGGINGFACEHUB_API_TOKEN"] = api_key
167
- st.session_state['rqa'][model] = init_qa(model)
 
168
 
169
  elif model == 'chatgpt-3.5-turbo':
170
  if 'OPENAI_API_KEY' not in os.environ:
171
  api_key = st.text_input('OpenAI API Key', type="password")
172
  st.markdown(
173
- "Get it for [Open AI](https://platform.openai.com/account/api-keys) or [Huggingface](https://huggingface.co/docs/hub/security-tokens)")
174
  else:
175
  api_key = os.environ['OPENAI_API_KEY']
176
 
177
  if api_key:
178
  # st.session_state['api_key'] = is_api_key_provided = True
179
- with st.spinner("Preparing environment"):
180
- st.session_state['api_keys']['chatgpt-3.5-turbo'] = api_key
181
- if 'OPENAI_API_KEY' not in os.environ:
182
- os.environ['OPENAI_API_KEY'] = api_key
183
- st.session_state['rqa'][model] = init_qa(model)
 
184
  # else:
185
  # is_api_key_provided = st.session_state['api_key']
186
 
187
  st.title("๐Ÿ“ Scientific Document Insight Q&A")
188
  st.subheader("Upload a scientific article in PDF, ask questions, get insights.")
189
 
 
 
190
  uploaded_file = st.file_uploader("Upload an article", type=("pdf", "txt"), on_change=new_file,
191
  disabled=st.session_state['model'] is not None and st.session_state['model'] not in
192
  st.session_state['api_keys'],
@@ -220,7 +237,7 @@ with st.sidebar:
220
  st.header("Documentation")
221
  st.markdown("https://github.com/lfoppiano/document-qa")
222
  st.markdown(
223
- """After entering your API Key (Open AI or Huggingface). Upload a scientific article as PDF document. You will see a spinner or loading indicator while the processing is in progress. Once the spinner stops, you can proceed to ask your questions.""")
224
 
225
  if st.session_state['git_rev'] != "unknown":
226
  st.markdown("**Revision number**: [" + st.session_state[
 
51
  if 'uploaded' not in st.session_state:
52
  st.session_state['uploaded'] = False
53
 
54
+ st.set_page_config(
55
+ page_title="Document Insights QA",
56
+ page_icon="๐Ÿ“",
57
+ initial_sidebar_state="expanded",
58
+ menu_items={
59
+ 'Get Help': 'https://github.com/lfoppiano/document-qa',
60
+ 'Report a bug': "https://github.com/lfoppiano/document-qa/issues",
61
+ 'About': "Upload a scientific article in PDF, ask questions, get insights."
62
+ }
63
+ )
64
+
65
+
66
  def new_file():
67
  st.session_state['loaded_embeddings'] = None
68
  st.session_state['doc_id'] = None
69
  st.session_state['uploaded'] = True
70
 
71
+
72
  # @st.cache_resource
73
  def init_qa(model):
74
  if model == 'chatgpt-3.5-turbo':
 
147
  # is_api_key_provided = st.session_state['api_key']
148
 
149
  with st.sidebar:
 
 
 
150
  st.session_state['model'] = model = st.radio(
151
  "Model",
152
  ("chatgpt-3.5-turbo", "mistral-7b-instruct-v0.1"), # , "llama-2-70b-chat"),
153
  index=1,
154
  captions=[
155
  "ChatGPT 3.5 Turbo + Ada-002-text (embeddings)",
156
+ "Mistral-7B-Instruct-V0.1 + Sentence BERT (embeddings) :free:"
157
+ # "LLama2-70B-Chat + Sentence BERT (embeddings) :free:",
158
  ],
159
  help="Select the LLM model and embeddings you want to use.",
160
  disabled=st.session_state['doc_id'] is not None or st.session_state['uploaded'])
161
 
162
+ st.markdown(
163
+ ":warning: Mistral is free to use, however requests might hit limits of the huggingface free API and fail. :warning: ")
164
+
165
  if model == 'mistral-7b-instruct-v0.1' or model == 'llama-2-70b-chat':
166
  if 'HUGGINGFACEHUB_API_TOKEN' not in os.environ:
167
  api_key = st.text_input('Huggingface API Key', type="password")
168
 
169
  st.markdown(
170
+ "Get it [here](https://huggingface.co/docs/hub/security-tokens)")
171
  else:
172
  api_key = os.environ['HUGGINGFACEHUB_API_TOKEN']
173
 
174
  if api_key:
175
  # st.session_state['api_key'] = is_api_key_provided = True
176
+ if model not in st.session_state['rqa'] or model not in st.session_state['api_keys']:
177
+ with st.spinner("Preparing environment"):
178
+ st.session_state['api_keys'][model] = api_key
179
+ if 'HUGGINGFACEHUB_API_TOKEN' not in os.environ:
180
+ os.environ["HUGGINGFACEHUB_API_TOKEN"] = api_key
181
+ st.session_state['rqa'][model] = init_qa(model)
182
 
183
  elif model == 'chatgpt-3.5-turbo':
184
  if 'OPENAI_API_KEY' not in os.environ:
185
  api_key = st.text_input('OpenAI API Key', type="password")
186
  st.markdown(
187
+ "Get it [here](https://platform.openai.com/account/api-keys)")
188
  else:
189
  api_key = os.environ['OPENAI_API_KEY']
190
 
191
  if api_key:
192
  # st.session_state['api_key'] = is_api_key_provided = True
193
+ if model not in st.session_state['rqa'] or model not in st.session_state['api_keys']:
194
+ with st.spinner("Preparing environment"):
195
+ st.session_state['api_keys'][model] = api_key
196
+ if 'OPENAI_API_KEY' not in os.environ:
197
+ os.environ['OPENAI_API_KEY'] = api_key
198
+ st.session_state['rqa'][model] = init_qa(model)
199
  # else:
200
  # is_api_key_provided = st.session_state['api_key']
201
 
202
  st.title("๐Ÿ“ Scientific Document Insight Q&A")
203
  st.subheader("Upload a scientific article in PDF, ask questions, get insights.")
204
 
205
+ st.markdown(":warning: Do not upload sensitive data. We **temporarily** store text from the uploaded PDF documents solely for the purpose of processing your request, and we **do not assume responsibility** for any subsequent use or handling of the data submitted to third parties LLMs.")
206
+
207
  uploaded_file = st.file_uploader("Upload an article", type=("pdf", "txt"), on_change=new_file,
208
  disabled=st.session_state['model'] is not None and st.session_state['model'] not in
209
  st.session_state['api_keys'],
 
237
  st.header("Documentation")
238
  st.markdown("https://github.com/lfoppiano/document-qa")
239
  st.markdown(
240
+ """Upload a scientific article as PDF document. Once the spinner stops, you can proceed to ask your questions.""")
241
 
242
  if st.session_state['git_rev'] != "unknown":
243
  st.markdown("**Revision number**: [" + st.session_state[