Create README.md
#1
by Elena0111 - opened
README.md
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from transformers import GPT2Tokenizer, GPT2LMHeadModel, Trainer, TrainingArguments
|
| 2 |
+
|
| 3 |
+
# 加載預訓練模型和標記器
|
| 4 |
+
model = GPT2LMHeadModel.from_pretrained('gpt2')
|
| 5 |
+
tokenizer = GPT2Tokenizer.from_pretrained('gpt2')
|
| 6 |
+
|
| 7 |
+
# 準備數據集(假設您有一個自定義的數據集)
|
| 8 |
+
train_dataset = https://www.ird.gov.hk/datagovhk/brfee.csv # 您的訓練數據集
|
| 9 |
+
|
| 10 |
+
# 設置訓練參數
|
| 11 |
+
training_args = TrainingArguments(
|
| 12 |
+
output_dir='./results',
|
| 13 |
+
num_train_epochs=3,
|
| 14 |
+
per_device_train_batch_size=4,
|
| 15 |
+
save_steps=10_000,
|
| 16 |
+
save_total_limit=2,
|
| 17 |
+
)
|
| 18 |
+
|
| 19 |
+
# 訓練模型
|
| 20 |
+
trainer = Trainer(
|
| 21 |
+
model=model,
|
| 22 |
+
args=training_args,
|
| 23 |
+
train_dataset=train_dataset,
|
| 24 |
+
)
|
| 25 |
+
|
| 26 |
+
trainer.train()
|