Update app.py

app.py

@@ -8,7 +8,7 @@ import torch
 from transformers import AutoTokenizer, AutoModelForCausalLM, TextIteratorStreamer
 
 
-model_id = "Qwen/Qwen2.5-32B-Instruct"
+model_id = "Qwen/Qwen2.5-14B-Instruct"
 assistant_id = "Qwen/Qwen2.5-0.5B-Instruct"
 
 model = AutoModelForCausalLM.from_pretrained(model_id, load_in_4bit=True, device_map="auto")
@@ -61,8 +61,8 @@ def reset_textbox():
 with gr.Blocks() as demo:
     gr.Markdown(
         "# 🤗 Assisted Generation Demo\n"
-        f"- Model: {model_id} (4-bit quant, ~16GB)\n"
-        f"- Assistant Model: {assistant_id} (FP16, ~1GB)\n"
+        f"- Model: {model_id} (4-bit quant, 14B params, GPU memory = ~7GB)\n"
+        f"- Assistant Model: {assistant_id} (FP16, 0.5B params, GPU memory = ~1GB)\n"
         "- Recipe for speedup: a) >10x model size difference in parameters; b) assistant trained similarly; c) CPU is not a bottleneck"
     )