fix: revert to non streaming

app.py

@@ -28,28 +28,36 @@ def inference(prompt: str, max_tokens: int = 256) -> str:
     )
 
     inputs = tokenizer(text, return_tensors="pt").to("cuda")
+
+    outputs = model.generate(
+        inputs,
+        max_new_tokens = 100, # Increase for longer outputs!
+        temperature = 0.7, top_p = 0.8, top_k = 20, # For non thinking
+        streamer = TextStreamer(tokenizer, skip_prompt = True),
+    )
+    return tokenizer.decode(outputs[0], skip_special_tokens=True)
     
-    # Use TextIteratorStreamer instead of TextStreamer
-    streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
+    # # Use TextIteratorStreamer instead of TextStreamer
+    # streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
     
-    generation_kwargs = dict(
-        **inputs,
-        max_new_tokens=max_tokens,
-        temperature=0.7,
-        top_p=0.8,
-        top_k=20,
-        streamer=streamer,
-    )
+    # generation_kwargs = dict(
+    #     **inputs,
+    #     max_new_tokens=max_tokens,
+    #     temperature=0.7,
+    #     top_p=0.8,
+    #     top_k=20,
+    #     streamer=streamer,
+    # )
     
-    # Run generation in separate thread
-    thread = Thread(target=model.generate, kwargs=generation_kwargs)
-    thread.start()
+    # # Run generation in separate thread
+    # thread = Thread(target=model.generate, kwargs=generation_kwargs)
+    # thread.start()
     
-    # Yield tokens as they come
-    generated_text = ""
-    for new_text in streamer:
-        generated_text += new_text
-        yield generated_text  # yield cumulative text for Gradio
+    # # Yield tokens as they come
+    # generated_text = ""
+    # for new_text in streamer:
+    #     generated_text += new_text
+    #     yield generated_text  # yield cumulative text for Gradio
 
 demo = gr.Interface(
     fn=inference,