returned classic generating

app.py

@@ -39,20 +39,21 @@ def load_model():
 
 
 @torch.inference_mode()
-def generate_stream(prompt, max_new_tokens=200, temperature=0.9, top_k=50):
-    """Generira tekst iz prompta i yield-a rezultat za streaming efekt bez ponavljanja."""
+def generate(prompt, max_new_tokens=200, temperature=0.9, top_k=50):
+    """Generiranje teksta iz prompta"""
     model = load_model()
     enc = state["enc"]
+
     x = torch.tensor([enc.encode(prompt)], dtype=torch.long, device=DEVICE)
+
     y = model.generate(
         x,
         max_new_tokens=int(max_new_tokens),
         temperature=float(temperature),
         top_k=int(top_k) if top_k > 0 else None
-    )[0].tolist()
-    # Stream output token by token
-    for i in range(len(x[0]), len(y)):
-        yield enc.decode(y[:i+1])
+    )
+
+    return enc.decode(y[0].tolist())
 
 # Gradio UI
 with gr.Blocks(title="🎬 Final Thesis Plot Generator") as demo:
@@ -63,9 +64,9 @@ with gr.Blocks(title="🎬 Final Thesis Plot Generator") as demo:
     temperature = gr.Slider(0.1, 1.5, value=0.9, step=0.1, label="Temperature")
     top_k = gr.Slider(0, 100, value=50, step=5, label="Top-K (0 = off)")
     btn = gr.Button("Generate")
+    output = gr.Textbox(label="Output", lines=15)
 
-    output = gr.Textbox(label="Output", lines=15, interactive=False)
-    btn.click(generate_stream, [prompt, max_new_tokens, temperature, top_k], output)
+    btn.click(generate, [prompt, max_new_tokens, temperature, top_k], output)
 
 if __name__ == "__main__":
     demo.launch()