Spaces:
Sleeping
Sleeping
Update augment_intents.py
Browse files- augment_intents.py +34 -34
augment_intents.py
CHANGED
|
@@ -1,34 +1,34 @@
|
|
| 1 |
-
import json
|
| 2 |
-
import random
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
prefixes = ["ممكن", "لو سمحت", "لو تكرمت", "أبغى", "ودي", "احتاج"]
|
| 6 |
-
suffixes = ["بالضبط", "من فضلك", "الله يعافيك", "بسرعة", "تكفى"]
|
| 7 |
-
|
| 8 |
-
def augment_patterns(patterns, n=3):
|
| 9 |
-
"""يزيد الأمثلة بإضافة prefixes و suffixes"""
|
| 10 |
-
augmented = []
|
| 11 |
-
for p in patterns:
|
| 12 |
-
augmented.append(p)
|
| 13 |
-
for _ in range(n):
|
| 14 |
-
new_p = p
|
| 15 |
-
if random.random() > 0.5:
|
| 16 |
-
new_p = random.choice(prefixes) + " " + new_p
|
| 17 |
-
if random.random() > 0.5:
|
| 18 |
-
new_p = new_p + " " + random.choice(suffixes)
|
| 19 |
-
augmented.append(new_p)
|
| 20 |
-
return list(set(augmented))
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
with open("intents.json", encoding="utf-8") as f:
|
| 24 |
-
data = json.load(f)
|
| 25 |
-
|
| 26 |
-
|
| 27 |
-
for intent in data["intents"]:
|
| 28 |
-
intent["patterns"] = augment_patterns(intent["patterns"], n=5)
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
with open("intents_augmented.json", "w", encoding="utf-8") as f:
|
| 32 |
-
json.dump(data, f, ensure_ascii=False, indent=2)
|
| 33 |
-
|
| 34 |
-
print("
|
|
|
|
| 1 |
+
import json
|
| 2 |
+
import random
|
| 3 |
+
|
| 4 |
+
|
| 5 |
+
prefixes = ["ممكن", "لو سمحت", "لو تكرمت", "أبغى", "ودي", "احتاج"]
|
| 6 |
+
suffixes = ["بالضبط", "من فضلك", "الله يعافيك", "بسرعة", "تكفى"]
|
| 7 |
+
|
| 8 |
+
def augment_patterns(patterns, n=3):
|
| 9 |
+
"""يزيد الأمثلة بإضافة prefixes و suffixes"""
|
| 10 |
+
augmented = []
|
| 11 |
+
for p in patterns:
|
| 12 |
+
augmented.append(p)
|
| 13 |
+
for _ in range(n):
|
| 14 |
+
new_p = p
|
| 15 |
+
if random.random() > 0.5:
|
| 16 |
+
new_p = random.choice(prefixes) + " " + new_p
|
| 17 |
+
if random.random() > 0.5:
|
| 18 |
+
new_p = new_p + " " + random.choice(suffixes)
|
| 19 |
+
augmented.append(new_p)
|
| 20 |
+
return list(set(augmented))
|
| 21 |
+
|
| 22 |
+
|
| 23 |
+
with open("intents.json", encoding="utf-8") as f:
|
| 24 |
+
data = json.load(f)
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
for intent in data["intents"]:
|
| 28 |
+
intent["patterns"] = augment_patterns(intent["patterns"], n=5)
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
with open("intents_augmented.json", "w", encoding="utf-8") as f:
|
| 32 |
+
json.dump(data, f, ensure_ascii=False, indent=2)
|
| 33 |
+
|
| 34 |
+
print("intents_augmented.json")
|