Spaces:
Sleeping
Sleeping
Delete utils.py
Browse files
utils.py
DELETED
@@ -1,28 +0,0 @@
|
|
1 |
-
from transformers import Trainer, TrainingArguments, DataCollatorForLanguageModeling
|
2 |
-
from datasets import Dataset
|
3 |
-
|
4 |
-
def preprocess_data(df, tokenizer):
|
5 |
-
df["text"] = df.apply(lambda row: f"Question: {row['Question']} Answer: {row['Answer']}", axis=1)
|
6 |
-
dataset = Dataset.from_pandas(df)
|
7 |
-
dataset = dataset.map(lambda x: tokenizer(x["text"], truncation=True, padding="max_length", max_length=512), batched=True)
|
8 |
-
return dataset
|
9 |
-
|
10 |
-
def train_model(model, tokenizer, dataset, output_dir):
|
11 |
-
training_args = TrainingArguments(
|
12 |
-
output_dir=output_dir,
|
13 |
-
per_device_train_batch_size=4,
|
14 |
-
num_train_epochs=1,
|
15 |
-
logging_dir="./logs",
|
16 |
-
save_steps=10,
|
17 |
-
logging_steps=10
|
18 |
-
)
|
19 |
-
data_collator = DataCollatorForLanguageModeling(tokenizer=tokenizer, mlm=False)
|
20 |
-
trainer = Trainer(
|
21 |
-
model=model,
|
22 |
-
args=training_args,
|
23 |
-
train_dataset=dataset,
|
24 |
-
data_collator=data_collator
|
25 |
-
)
|
26 |
-
trainer.train()
|
27 |
-
model.save_pretrained(output_dir)
|
28 |
-
tokenizer.save_pretrained(output_dir)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|