Deploy DeepSeek LLM chatbot

app.py

@@ -0,0 +1,66 @@
+import gradio as gr
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+
+# Load DeepSeek LLM
+model_name = "deepseek-ai/deepseek-llm-7b-chat"  
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.float16, device_map="auto")
+
+SYSTEM_PROMPT = "You are a helpful AI assistant. Keep responses concise and informative."
+
+def generate_response(message, history):
+    history = history or []
+    history.append(("User", message))
+    
+    # Add system message for better guidance
+    input_text = f"[SYSTEM] {SYSTEM_PROMPT}\n" + tokenizer.apply_chat_template(history, tokenize=False, add_generation_prompt=True)
+    inputs = tokenizer(input_text, return_tensors="pt").to("cuda")
+
+    streamer = tokenizer.streamer()
+    model.generate(**inputs, streamer=streamer, max_length=512, pad_token_id=tokenizer.eos_token_id)
+    
+    bot_message = ""
+    for token in streamer:
+        bot_message += token
+        yield bot_message  
+
+# Create Gradio Chatbot UI with streaming
+with gr.Blocks() as demo:
+    gr.Markdown("### 🚀 DeepSeek LLM Chatbot (Streaming & Improved UI)")
+    
+    chatbot = gr.Chatbot()
+    msg = gr.Textbox(placeholder="Type your message here...", label="Your Message")
+    clear_btn = gr.Button("Clear Chat")
+    
+    def respond(message, history):
+        history = history or []
+        bot_response = generate_response(message, history)
+        return bot_response, history + [("User", message), ("Bot", bot_response)]
+    
+    msg.submit(respond, inputs=[msg, chatbot], outputs=[chatbot, msg])
+    clear_btn.click(lambda: ([], ""), outputs=[chatbot, msg])
+
+demo.launch()
+
+
+
+# with gr.Blocks() as demo:
+#     gr.Markdown("### 🚀 DeepSeek LLM Chatbot (Streaming Enabled)")
+#     chat = gr.ChatInterface(fn=generate_response)
+    
+# demo.launch()
+
+'''
+✅ Uses streaming (streamer=tokenizer.streamer())
+✅ Returns tokens in real-time instead of waiting for full response
+✅ Improved UI with gr.Blocks()
+
+✅ System prompt ensures responses are concise & helpful
+✅ Chat history is structured more clearly
+
+✅ Retains chat history
+✅ "Clear Chat" button
+✅ Better UI layout with Markdown & structured input boxes
+
+'''

requirements.txt

@@ -0,0 +1,4 @@
+gradio
+transformers
+torch
+accelerate