Spaces:
Running
Running
aliasgerovs
commited on
Commit
•
c10af95
1
Parent(s):
42fb43c
Update app.py
Browse files
app.py
CHANGED
@@ -224,9 +224,11 @@ def split_text_allow_complete_sentences_nltk(text, max_length=256, tolerance=30,
|
|
224 |
|
225 |
if type_det == 'bc':
|
226 |
tokenizer = text_bc_tokenizer
|
|
|
227 |
|
228 |
elif type_det == 'mc':
|
229 |
tokenizer = text_mc_tokenizer
|
|
|
230 |
|
231 |
for sentence in sentences:
|
232 |
tokens = tokenizer.tokenize(sentence)
|
@@ -282,7 +284,7 @@ def predict_bc(model, tokenizer, text):
|
|
282 |
with torch.no_grad():
|
283 |
model.eval()
|
284 |
tokens = text_bc_tokenizer(
|
285 |
-
text, padding='max_length', truncation=True, max_length=
|
286 |
).to(device)
|
287 |
output = model(**tokens)
|
288 |
output_norm = softmax(output.logits.detach().cpu().numpy(), 1)[0]
|
|
|
224 |
|
225 |
if type_det == 'bc':
|
226 |
tokenizer = text_bc_tokenizer
|
227 |
+
max_length = 333
|
228 |
|
229 |
elif type_det == 'mc':
|
230 |
tokenizer = text_mc_tokenizer
|
231 |
+
max_length = 256
|
232 |
|
233 |
for sentence in sentences:
|
234 |
tokens = tokenizer.tokenize(sentence)
|
|
|
284 |
with torch.no_grad():
|
285 |
model.eval()
|
286 |
tokens = text_bc_tokenizer(
|
287 |
+
text, padding='max_length', truncation=True, max_length=333, return_tensors="pt"
|
288 |
).to(device)
|
289 |
output = model(**tokens)
|
290 |
output_norm = softmax(output.logits.detach().cpu().numpy(), 1)[0]
|