Upload app.py with huggingface_hub

app.py

@@ -0,0 +1,59 @@
+import gradio as gr
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+# Load model and tokenizer
+model_id = "LiquidAI/LFM2-2.6B"
+model = AutoModelForCausalLM.from_pretrained(
+    model_id,
+    device_map="auto",
+    torch_dtype="bfloat16",
+)
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+
+def chat_with_model(message, history):
+    # Format conversation history
+    conversation = []
+    for user_msg, assistant_msg in history:
+        conversation.append({"role": "user", "content": user_msg})
+        conversation.append({"role": "assistant", "content": assistant_msg})
+    conversation.append({"role": "user", "content": message})
+    
+    # Apply chat template
+    input_ids = tokenizer.apply_chat_template(
+        conversation,
+        add_generation_prompt=True,
+        return_tensors="pt",
+        tokenize=True,
+    ).to(model.device)
+    
+    # Generate response
+    output = model.generate(
+        input_ids,
+        do_sample=True,
+        temperature=0.3,
+        min_p=0.15,
+        repetition_penalty=1.05,
+        max_new_tokens=512,
+        pad_token_id=tokenizer.eos_token_id,
+    )
+    
+    # Decode only the newly generated tokens, skipping the prompt
+    response = tokenizer.decode(output[0][input_ids.shape[-1]:], skip_special_tokens=True)
+    
+    return response
+
+# Create Gradio interface
+iface = gr.ChatInterface(
+    fn=chat_with_model,
+    title="LFM2-2.6B Chatbot",
+    description="A chatbot powered by LiquidAI/LFM2-2.6B. Built with [anycoder](https://huggingface.co/spaces/akhaliq/anycoder).",
+    theme="soft",
+    examples=[
+        ["What is C. elegans?"],
+        ["Write a short story about a robot who discovers music."],
+        ["Explain the importance of the transformer architecture in NLP."],
+    ],
+)
+
+iface.launch()