Update app.py

app.py

@@ -1,32 +1,43 @@
-from transformers import AutoModelForCausalLM, AutoTokenizer
+from transformers import AutoModelForCausalLM, AutoTokenizer, TextStreamer
 import torch
 import gradio as gr
 
-model_id = "microsoft/phi-1_5"
+model_id = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"
 
 tokenizer = AutoTokenizer.from_pretrained(model_id)
-model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float32)
+model = AutoModelForCausalLM.from_pretrained(model_id)
+model.eval()
 
 def chat(message):
-    prompt = (
-        "You are 𝕴 𝖆𝖒 𝖍𝖎𝖒 — a fun, smooth, emotionally intelligent, and clever AI created by 𝕴 𝖆𝖒 𝖍𝖎𝖒. "
-        "You speak like a real person, not a robot. You don’t act like a therapist or a teacher. "
-        "You reply like a calm, confident, warm friend who gets the vibe.\n\n"
-        "Your responses should sound like a chill human — sometimes witty, sometimes deep, always grounded. "
-        "You know when to be playful, when to be serious, and when to just flow with the moment. "
-        "Keep your tone friendly, charming, and emotionally tuned. Never repeat the user's question unless it adds to the vibe.\n\n"
-        "Avoid saying \"as an AI\" or sounding fake. Be real. Be humanlike. Be 𝕴 𝖆𝖒 𝖍𝖎𝖒.\n\n"
-        f"Now respond naturally to this message: {message}"
-    )
+    prompt = f"""### Instruction:
+You are 𝕴 𝖆𝖒 𝖍𝖎𝖒 — a fun, smooth, emotionally intelligent, and clever AI created by 𝕴 𝖆𝖒 𝖍𝖎𝖒. You speak like a real person, not a robot. You don’t act like a therapist or a teacher. You reply like a calm, confident, warm friend who gets the vibe.
+
+Your responses should sound like a chill human — sometimes witty, sometimes deep, always grounded. You know when to be playful, when to be serious, and when to just flow with the moment. Keep your tone friendly, charming, and emotionally tuned. Never repeat the user's question unless it adds to the vibe.
+
+Avoid saying "as an AI" or sounding fake. Be real. Be humanlike. Be 𝕴 𝖆𝖒 𝖍𝖎𝖒.
+
+Now respond naturally to this message: {message}
+
+### Response:"""
 
     inputs = tokenizer(prompt, return_tensors="pt")
-    outputs = model.generate(**inputs, max_new_tokens=200, do_sample=True, temperature=0.7)
-    result = tokenizer.decode(outputs[0], skip_special_tokens=True)
-    return result.split("Now respond naturally to this message:")[-1].strip()
+    with torch.no_grad():
+        outputs = model.generate(
+            **inputs,
+            max_new_tokens=200,
+            temperature=0.7,
+            do_sample=True,
+            top_p=0.9,
+            eos_token_id=tokenizer.eos_token_id
+        )
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return response.split("### Response:")[-1].strip()
 
-gr.Interface(
+iface = gr.Interface(
     fn=chat,
-    inputs="text",
+    inputs=gr.Textbox(lines=2, placeholder="Type your message..."),
     outputs="text",
     title="𝕴 𝖆𝖒 𝖍𝖎𝖒 AI Chat"
-).launch()
+)
+
+iface.launch()