Create app.py

app.py

@@ -0,0 +1,53 @@
+import gradio as gr
+import torch
+import torch.nn.functional as F
+from transformers import AutoTokenizer, pipeline
+from optimum.onnxruntime import ORTModelForCausalLM
+
+# Load your custom tokenizer and ONNX model
+# Make sure the tokenizer and model files are in the same repo
+tokenizer = AutoTokenizer.from_pretrained("Smilyai-labs/Sam-3.0-2-onnx")
+model = ORTModelForCausalLM.from_pretrained("Smilyai-labs/Sam-3.0-2-onnx")
+
+# Define a function to generate text
+def generate_text(prompt, max_length=128, temperature=0.8, top_k=60, top_p=0.9):
+    # Create a text generation pipeline
+    gen_pipeline = pipeline(
+        "text-generation",
+        model=model,
+        tokenizer=tokenizer,
+        device="cpu"  # ONNX models run on CPU by default
+    )
+    
+    # Use the pipeline to generate text
+    generated_text = gen_pipeline(
+        prompt,
+        max_length=max_length,
+        temperature=temperature,
+        top_k=top_k,
+        top_p=top_p,
+        do_sample=True,
+    )
+    
+    # Return the generated text, cleaning up any special tokens
+    return generated_text[0]["generated_text"]
+
+# Create a Gradio interface
+# The fn parameter points to your generation function
+# The inputs and outputs define the UI and API endpoint
+demo = gr.Interface(
+    fn=generate_text,
+    inputs=[
+        gr.Textbox(label="Prompt", lines=2),
+        gr.Slider(minimum=10, maximum=512, value=128, label="Max Length"),
+        gr.Slider(minimum=0.1, maximum=2.0, value=0.8, label="Temperature"),
+        gr.Slider(minimum=1, maximum=100, value=60, label="Top K"),
+        gr.Slider(minimum=0.1, maximum=1.0, value=0.9, label="Top P"),
+    ],
+    outputs="text",
+    title="SmilyAI Sam 3.0-2 ONNX Text Generation",
+    description="A simple API and UI for text generation using the ONNX version of Sam 3.0-2."
+)
+
+# Launch the Gradio app
+demo.launch()