Add Qwen3.5-35B-A3B AWQ chat app with ZeroGPU support

app.py

@@ -1,69 +1,133 @@
+import spaces
 import gradio as gr
-from huggingface_hub import InferenceClient
-
-
-def respond(
-    message,
-    history: list[dict[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-    hf_token: gr.OAuthToken,
-):
-    """
-    For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-    """
-    client = InferenceClient(token=hf_token.token, model="openai/gpt-oss-20b")
-
-    messages = [{"role": "system", "content": system_message}]
-
-    messages.extend(history)
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        choices = message.choices
-        token = ""
-        if len(choices) and choices[0].delta.content:
-            token = choices[0].delta.content
-
-        response += token
-        yield response
-
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-chatbot = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
+import torch
+from transformers import AutoTokenizer, AutoModelForCausalLM, TextIteratorStreamer
+from threading import Thread
+
+MODEL_ID = "QuantTrio/Qwen3.5-35B-A3B-AWQ"
+
+print("Loading tokenizer...")
+tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, trust_remote_code=True)
+
+print("Loading model...")
+model = AutoModelForCausalLM.from_pretrained(
+        MODEL_ID,
+        torch_dtype=torch.float16,
+        device_map="cpu",
+        trust_remote_code=True,
 )
+model.eval()
+print("Model loaded.")
+
+@spaces.GPU(duration=120)
+def respond(message, history, system_message, max_new_tokens, temperature, top_p, enable_thinking):
+        model.to("cuda")
+
+        messages = [{"role": "system", "content": system_message}]
+        for user_msg, assistant_msg in history:
+                    if user_msg:
+                                    messages.append({"role": "user", "content": user_msg})
+                                if assistant_msg:
+                                    messages.append({"role": "assistant", "content": assistant_msg})
+                            messages.append({"role": "user", "content": message})
+
+                text = tokenizer.apply_chat_template(
+                            messages,
+                            tokenize=False,
+                            add_generation_prompt=True,
+                            chat_template_kwargs={"enable_thinking": enable_thinking},
+                )
+    inputs = tokenizer([text], return_tensors="pt").to("cuda")
+
+    streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
+
+    generation_kwargs = dict(
+                **inputs,
+                streamer=streamer,
+                max_new_tokens=max_new_tokens,
+                temperature=temperature,
+                top_p=top_p,
+                top_k=20,
+                do_sample=True,
+                repetition_penalty=1.1,
+    )
+
+    thread = Thread(target=model.generate, kwargs=generation_kwargs)
+    thread.start()
+
+    partial = ""
+    thinking_done = False
+    display_text = ""
+
+    for new_text in streamer:
+                partial += new_text
+                if enable_thinking:
+                                if "</think>" in partial and not thinking_done:
+                                                    thinking_done = True
+                                                    parts = partial.split("</think>", 1)
+                                                    display_text = parts[1].lstrip("\n")
+                                elif thinking_done:
+                                                    display_text = partial.split("</think>", 1)[1].lstrip("\n")
+                                else:
+                                                    display_text = ""
+                else:
+                                display_text = partial
+                            yield display_text
+
+    model.to("cpu")
+    torch.cuda.empty_cache()
+
+
+with gr.Blocks(title="Qwen3.5-35B-A3B Chat") as demo:
+        gr.Markdown("# Qwen3.5-35B-A3B AWQ Chat\nPowered by ZeroGPU (H200) | 4-bit AWQ quantized | 25.5 GB")
+
+    chatbot = gr.Chatbot(height=500, label="Qwen3.5-35B-A3B")
+
+    with gr.Row():
+                msg = gr.Textbox(
+                    placeholder="Type your message here...",
+                    label="Your message",
+                    scale=4,
+    )
+        send_btn = gr.Button("Send", variant="primary", scale=1)
+
+    with gr.Accordion("Settings", open=False):
+                system_message = gr.Textbox(
+                    value="You are a helpful, smart, and concise AI assistant.",
+                    label="System message",
+    )
+        enable_thinking = gr.Checkbox(value=True, label="Enable thinking mode (chain of thought)")
+        max_new_tokens = gr.Slider(minimum=64, maximum=8192, value=2048, step=64, label="Max new tokens")
+        temperature = gr.Slider(minimum=0.1, maximum=2.0, value=1.0, step=0.1, label="Temperature")
+        top_p = gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p")
+
+    def user_turn(user_message, history):
+                return "", history + [[user_message, None]]
+
+            def bot_turn(history, system_message, max_new_tokens, temperature, top_p, enable_thinking):
+                        user_message = history[-1][0]
+                        bot_history = history[:-1]
+                        response = ""
+                        for chunk in respond(user_message, bot_history, system_message, max_new_tokens, temperature, top_p, enable_thinking):
+                                        response = chunk
+                                        history[-1][1] = response
+                                        yield history
 
-with gr.Blocks() as demo:
-    with gr.Sidebar():
-        gr.LoginButton()
-    chatbot.render()
+                                msg.submit(user_turn, [msg, chatbot], [msg, chatbot]).then(
+                                            bot_turn, [chatbot, system_message, max_new_tokens, temperature, top_p, enable_thinking], chatbot
+                                )
+    send_btn.click(user_turn, [msg, chatbot], [msg, chatbot]).then(
+                bot_turn, [chatbot, system_message, max_new_tokens, temperature, top_p, enable_thinking], chatbot
+    )
 
+    gr.Examples(
+                examples=[
+                                ["What is the square root of 144? Think step by step."],
+                                ["Write a Python function to check if a number is prime."],
+                                ["Explain quantum entanglement in simple terms."],
+                                ["Translate 'Hello, how are you?' into French, Spanish, and Japanese."],
+                ],
+                inputs=msg,
+    )
 
-if __name__ == "__main__":
-    demo.launch()
+demo.launch()