Spaces:
Build error
Build error
Update app.py
Browse files
app.py
CHANGED
@@ -107,13 +107,8 @@ def transcribe(microphone, file_upload, batch_size=15):
|
|
107 |
|
108 |
return warn_output + text.strip(), system_info
|
109 |
|
110 |
-
|
111 |
-
|
112 |
-
|
113 |
-
|
114 |
# ------------summary section------------
|
115 |
|
116 |
-
|
117 |
# ------------for app integration later------------
|
118 |
|
119 |
nlp = spacy.blank("nb") # codename 'nb' = Norwegian Bokmål
|
@@ -145,9 +140,9 @@ def preprocess_text(text, file_upload):
|
|
145 |
|
146 |
@spaces.GPU()
|
147 |
def summarize_text(text, file_upload):
|
148 |
-
#
|
149 |
-
# ----add same if/elif logic as above here----
|
150 |
-
#
|
151 |
preprocessed_text = preprocess_text(text)
|
152 |
inputs = summarization_model(preprocessed_text, max_length=1024, return_tensors="pt", truncation=True)
|
153 |
inputs = inputs.to(device)
|
@@ -167,9 +162,9 @@ def build_similarity_matrix(sentences):
|
|
167 |
# PageRank
|
168 |
@spaces.GPU()
|
169 |
def graph_based_summary(text, file_upload, num_paragraphs=3):
|
170 |
-
#
|
171 |
-
# ----add same if/elif logic as above here----
|
172 |
-
#
|
173 |
sentences = [sent.text for sent in doc.sents]
|
174 |
if len(sentences) < num_paragraphs:
|
175 |
return ' '.join(sentences)
|
@@ -211,7 +206,7 @@ def lex_rank_summary(text, file_upload, num_paragraphs=3, threshold=0.1):
|
|
211 |
@spaces.GPU()
|
212 |
def text_rank_summary(text, file_upload, num_paragraphs=3):
|
213 |
|
214 |
-
if (text is not None) and (file_upload is None):
|
215 |
doc = nlp(text)
|
216 |
|
217 |
elif (text is None) and (file_upload is not None):
|
@@ -235,9 +230,9 @@ def save_to_pdf(text, summary):
|
|
235 |
pdf = FPDF()
|
236 |
pdf.add_page()
|
237 |
pdf.set_font("Arial", size=12)
|
238 |
-
#
|
239 |
-
# ----add same if/elif logic as above here----
|
240 |
-
#
|
241 |
if text:
|
242 |
pdf.multi_cell(0, 10, "Text:\n" + text)
|
243 |
|
@@ -258,21 +253,20 @@ with iface:
|
|
258 |
gr.Markdown(HEADER_INFO)
|
259 |
|
260 |
with gr.Row():
|
261 |
-
|
262 |
-
|
263 |
-
|
264 |
-
|
265 |
-
|
266 |
-
|
267 |
|
268 |
-
|
269 |
-
|
270 |
-
|
271 |
|
272 |
-
|
273 |
-
|
274 |
-
|
275 |
-
|
276 |
|
277 |
with gr.Tabs():
|
278 |
|
@@ -294,7 +288,6 @@ with iface:
|
|
294 |
summarize_uploaded_button_graph.click(fn=graph_based_summary(file_upload), inputs=[text_input_graph], outputs=[summary_output_graph])
|
295 |
|
296 |
with gr.TabItem("Summary | LexRank"):
|
297 |
-
with gr.Blocks():
|
298 |
text_output = gr.Textbox(label="Transcription Output")
|
299 |
text_input_lex = gr.Textbox(label="Input Text", placeholder="txt2summarize")
|
300 |
summary_output_lex = gr.Textbox(label="LexRank | cosine similarity")
|
@@ -321,7 +314,6 @@ with iface:
|
|
321 |
""")
|
322 |
gr.Markdown("""
|
323 |
*Bjørn*: **sammendrag basert på i de setningene som ligner mest på hverandre fra teksten**
|
324 |
-
|
325 |
""")
|
326 |
|
327 |
summarize_transcribed_button_text_rank = gr.Button("Summary of Transcribed Text, Click Here")
|
@@ -339,4 +331,4 @@ with iface:
|
|
339 |
|
340 |
pdf_text_only.click(fn=lambda text: save_to_pdf(text, ""), inputs=[transcribed_text], outputs=[pdf_output])
|
341 |
pdf_summary_only.click(fn=lambda summary: save_to_pdf("", summary), inputs=[summary_output_graph, summary_output_lex, summary_output_text_rank], outputs=[pdf_output]) # Includes all summary outputs
|
342 |
-
pdf_both.click(fn=lambda text, summary: save_to_pdf(text, summary), inputs=[transcribed_text, summary_output_graph], outputs=[pdf_output])
|
|
|
107 |
|
108 |
return warn_output + text.strip(), system_info
|
109 |
|
|
|
|
|
|
|
|
|
110 |
# ------------summary section------------
|
111 |
|
|
|
112 |
# ------------for app integration later------------
|
113 |
|
114 |
nlp = spacy.blank("nb") # codename 'nb' = Norwegian Bokmål
|
|
|
140 |
|
141 |
@spaces.GPU()
|
142 |
def summarize_text(text, file_upload):
|
143 |
+
#
|
144 |
+
# ----add same if/elif logic as above here----
|
145 |
+
#
|
146 |
preprocessed_text = preprocess_text(text)
|
147 |
inputs = summarization_model(preprocessed_text, max_length=1024, return_tensors="pt", truncation=True)
|
148 |
inputs = inputs.to(device)
|
|
|
162 |
# PageRank
|
163 |
@spaces.GPU()
|
164 |
def graph_based_summary(text, file_upload, num_paragraphs=3):
|
165 |
+
#
|
166 |
+
# ----add same if/elif logic as above here----
|
167 |
+
#
|
168 |
sentences = [sent.text for sent in doc.sents]
|
169 |
if len(sentences) < num_paragraphs:
|
170 |
return ' '.join(sentences)
|
|
|
206 |
@spaces.GPU()
|
207 |
def text_rank_summary(text, file_upload, num_paragraphs=3):
|
208 |
|
209 |
+
if (text is not None) and (file_upload is not None):
|
210 |
doc = nlp(text)
|
211 |
|
212 |
elif (text is None) and (file_upload is not None):
|
|
|
230 |
pdf = FPDF()
|
231 |
pdf.add_page()
|
232 |
pdf.set_font("Arial", size=12)
|
233 |
+
#
|
234 |
+
# ----add same if/elif logic as above here----
|
235 |
+
#
|
236 |
if text:
|
237 |
pdf.multi_cell(0, 10, "Text:\n" + text)
|
238 |
|
|
|
253 |
gr.Markdown(HEADER_INFO)
|
254 |
|
255 |
with gr.Row():
|
256 |
+
gr.Markdown('''
|
257 |
+
##### Here you will get transcription output
|
258 |
+
##### ''')
|
259 |
+
microphone = gr.Audio(sources="microphone", type="filepath")
|
260 |
+
upload = gr.Audio(sources="upload", type="filepath")
|
|
|
261 |
|
262 |
+
transcribe_btn = gr.Button("Transcribe Interview")
|
263 |
+
text_output = gr.Textbox()
|
264 |
+
system_info = gr.Textbox(label="System Info")
|
265 |
|
266 |
+
transcribe_btn.click(transcribe_audio,
|
267 |
+
[microphone, upload], [text_output]
|
268 |
+
[system_info]
|
269 |
+
)
|
270 |
|
271 |
with gr.Tabs():
|
272 |
|
|
|
288 |
summarize_uploaded_button_graph.click(fn=graph_based_summary(file_upload), inputs=[text_input_graph], outputs=[summary_output_graph])
|
289 |
|
290 |
with gr.TabItem("Summary | LexRank"):
|
|
|
291 |
text_output = gr.Textbox(label="Transcription Output")
|
292 |
text_input_lex = gr.Textbox(label="Input Text", placeholder="txt2summarize")
|
293 |
summary_output_lex = gr.Textbox(label="LexRank | cosine similarity")
|
|
|
314 |
""")
|
315 |
gr.Markdown("""
|
316 |
*Bjørn*: **sammendrag basert på i de setningene som ligner mest på hverandre fra teksten**
|
|
|
317 |
""")
|
318 |
|
319 |
summarize_transcribed_button_text_rank = gr.Button("Summary of Transcribed Text, Click Here")
|
|
|
331 |
|
332 |
pdf_text_only.click(fn=lambda text: save_to_pdf(text, ""), inputs=[transcribed_text], outputs=[pdf_output])
|
333 |
pdf_summary_only.click(fn=lambda summary: save_to_pdf("", summary), inputs=[summary_output_graph, summary_output_lex, summary_output_text_rank], outputs=[pdf_output]) # Includes all summary outputs
|
334 |
+
pdf_both.click(fn=lambda text, summary: save_to_pdf(text, summary), inputs=[transcribed_text, summary_output_graph], outputs=[pdf_output])
|