Update app.py

app.py

@@ -1,12 +1,3 @@
-import gradio as gr
-from transformers import pipeline, AutoTokenizer
-
-gr.load("models/rinna/japanese-gpt2-medium").launch()
-
-
-# 日本語モデルを指定
-model_name = "rinna/japanese-gpt2-medium"
-
 from spaces import GPU
 
 @GPU(duration=120)
@@ -14,10 +5,13 @@ def generate_text(prompt, max_length):
     result = generator(prompt, max_length=max_length, num_return_sequences=1)
     return result[0]['generated_text']
 
+# 日本語モデルを指定
+model_name = "rinna/japanese-gpt2-medium"
 
 # トークナイザーとパイプラインの設定
 tokenizer = AutoTokenizer.from_pretrained(model_name)
-generator = pipeline('text-generation', model=model_name, tokenizer=tokenizer, device=0)  # device=0はGPUを使用する設定
+device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+generator = pipeline('text-generation', model=model_name, tokenizer=tokenizer, device=0 if torch.cuda.is_available() else -1)
 
 def generate_text(prompt, max_length):
     result = generator(prompt, max_length=max_length, num_return_sequences=1)
@@ -32,4 +26,4 @@ iface = gr.Interface(
     outputs=gr.Textbox(label="生成されたテキスト")
 )
 
-iface.launch()
+iface.launch(share=True)