Create app.py

app.py

@@ -0,0 +1,34 @@
+import gradio as gr
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+
+# Load your model and tokenizer WITHOUT 4-bit quantization
+model_name = "mherrador/CE5.0_expert"  
+device = torch.device("cpu") 
+
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    # No quantization_config here 
+    trust_remote_code=True,
+).to(device)  
+
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+
+# Function to generate recommendations
+def generate_recommendations(input_text):
+    inputs = tokenizer(input_text, return_tensors="pt").to(device)  # Move input to device
+    outputs = model.generate(**inputs, max_new_tokens=128)
+    recommendations = tokenizer.batch_decode(outputs)[0]
+    return recommendations
+
+# Create the Gradio interface
+iface = gr.Interface(
+    fn=generate_recommendations,
+    inputs=gr.Textbox(lines=5, placeholder="Enter your questions here..."),
+    outputs=gr.Textbox(lines=10),
+    title="Circular Economy Recommender",
+    description="Enter your questions about circular economy practices to get recommendations.",
+)
+
+# Launch the interface
+iface.launch()