Sebasrr2903
commited on
Commit
·
4e0ccbe
1
Parent(s):
9b04e3b
v.4
Browse files- orderIA.py +4 -1
orderIA.py
CHANGED
|
@@ -1,7 +1,10 @@
|
|
| 1 |
from transformers import AutoTokenizer, AutoModelForSequenceClassification, Trainer, TrainingArguments
|
| 2 |
import pandas as pd
|
| 3 |
from datasets import Dataset
|
|
|
|
|
|
|
| 4 |
|
|
|
|
| 5 |
# Cargar el dataset
|
| 6 |
data = [
|
| 7 |
{"text": "¿Qué hamburguesas tienen?", "label": 0},
|
|
@@ -14,7 +17,7 @@ df = pd.DataFrame(data)
|
|
| 14 |
dataset = Dataset.from_pandas(df)
|
| 15 |
|
| 16 |
# Tokenizar los datos
|
| 17 |
-
tokenizer = AutoTokenizer.from_pretrained("bert-base-multilingual-cased")
|
| 18 |
|
| 19 |
def preprocess_function(examples):
|
| 20 |
return tokenizer(examples['text'], truncation=True, padding=True)
|
|
|
|
| 1 |
from transformers import AutoTokenizer, AutoModelForSequenceClassification, Trainer, TrainingArguments
|
| 2 |
import pandas as pd
|
| 3 |
from datasets import Dataset
|
| 4 |
+
from transformers import BertTokenizerFast
|
| 5 |
+
# Carga del tokenizador
|
| 6 |
|
| 7 |
+
tokenizer = BertTokenizerFast.from_pretrained('Sebastian2903/SMARTORDERIA')
|
| 8 |
# Cargar el dataset
|
| 9 |
data = [
|
| 10 |
{"text": "¿Qué hamburguesas tienen?", "label": 0},
|
|
|
|
| 17 |
dataset = Dataset.from_pandas(df)
|
| 18 |
|
| 19 |
# Tokenizar los datos
|
| 20 |
+
#tokenizer = AutoTokenizer.from_pretrained("bert-base-multilingual-cased")
|
| 21 |
|
| 22 |
def preprocess_function(examples):
|
| 23 |
return tokenizer(examples['text'], truncation=True, padding=True)
|