Update app.py

app.py

@@ -1,3 +1,4 @@
+# Imports
 import gradio as gr
 import os
 import random
@@ -12,7 +13,7 @@ HF_TOKEN = os.environ.get("HF_TOKEN")
 REPO = "HuggingFaceTB/SmolLM2-1.7B-Instruct-GGUF"
 FILE = "smollm2-1.7b-instruct-q4_k_m.gguf"
 
-TIMEOUT = 30
+TIMEOUT = 60
 
 MAX_SEED = 9007199254740991
 
@@ -68,13 +69,17 @@ def generate(prompt, temperature, top_p, top_k, repetition_penalty, max_tokens,
     finally:
         timer.cancel()
 
+@spaces.GPU(duration=15)
+def gpu():
+    return
+    
 # Initialize
 model_base = "Any"
 model_quant = "Any Quant"
 
 with gr.Blocks() as demo:
-    gr.Markdown("# 👁️‍🗨️ WizardLM")
-    gr.Markdown("• ⚡ A text generation inference for one of the best open-source text models: WizardLM-2-8x22B.")
+    gr.Markdown("# 👁️‍🗨️ LM")
+    gr.Markdown("• ⚡ A text generation inference for any quant models.")
     gr.Markdown("• ⚠️ **WARNING!** The inference is very slow due to the model being **HUGE**; it takes about 10 seconds before it starts generating. Please avoid high max token parameters and sending large amounts of text. Note it uses CPU because running it on GPU overloads the model.")
     gr.Markdown(f"• 🔗 Link to models: [{model_base}]({model_base}) (BASE), [{model_quant}]({model_quant}) (QUANT)")