Spaces:
Running
Running
aliasgerovs
commited on
Commit
·
617df1d
1
Parent(s):
6aa7315
Update app.py
Browse files
app.py
CHANGED
@@ -175,7 +175,7 @@ AI DETECTION SECTION
|
|
175 |
"""
|
176 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
177 |
|
178 |
-
text_bc_model_path = "polygraf-ai/
|
179 |
text_bc_tokenizer = AutoTokenizer.from_pretrained(text_bc_model_path)
|
180 |
text_bc_model = AutoModelForSequenceClassification.from_pretrained(text_bc_model_path).to(device)
|
181 |
|
@@ -192,8 +192,9 @@ def update_character_count(text):
|
|
192 |
|
193 |
def predict_bc(model, tokenizer, text):
|
194 |
tokens = tokenizer(
|
195 |
-
text, padding=
|
196 |
).to(device)["input_ids"]
|
|
|
197 |
output = model(tokens)
|
198 |
output_norm = softmax(output.logits.detach().cpu().numpy(), 1)[0]
|
199 |
print("BC Score: ", output_norm)
|
|
|
175 |
"""
|
176 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
177 |
|
178 |
+
text_bc_model_path = "polygraf-ai/v3-bert-3-2m-trun-bc"
|
179 |
text_bc_tokenizer = AutoTokenizer.from_pretrained(text_bc_model_path)
|
180 |
text_bc_model = AutoModelForSequenceClassification.from_pretrained(text_bc_model_path).to(device)
|
181 |
|
|
|
192 |
|
193 |
def predict_bc(model, tokenizer, text):
|
194 |
tokens = tokenizer(
|
195 |
+
text, padding='max_length', truncation=True, max_length=256, return_tensors="pt"
|
196 |
).to(device)["input_ids"]
|
197 |
+
|
198 |
output = model(tokens)
|
199 |
output_norm = softmax(output.logits.detach().cpu().numpy(), 1)[0]
|
200 |
print("BC Score: ", output_norm)
|