Update app.py

app.py

@@ -1,101 +1,10 @@
 import gradio as gr
-from transformers import AutoModelForCausalLM, AutoTokenizer
-import torch
-import os
 
-# Enable faster downloads with hf_transfer
-os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"
-
-# Retrieve token from environment (set in Space secrets if gated)
-HF_TOKEN = os.getenv("HF_TOKEN")
-if not HF_TOKEN:
-    print("Warning: HF_TOKEN not found. If the model is gated, it will fail to load.")
-
-# Model repository ID (replace with your specific DeepscaleR model if different)
-MODEL_ID = "agentica-org/DeepScaleR-1.5B-Preview"
-
-# Load model and tokenizer
-def load_model():
-    try:
-        tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, token=HF_TOKEN)
-        model = AutoModelForCausalLM.from_pretrained(MODEL_ID, token=HF_TOKEN)
-        device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-        model.to(device)
-        return tokenizer, model, device
-    except Exception as e:
-        return None, None, str(e)
-
-# Load model at startup
-tokenizer, model, load_error = load_model()
-
-# Inference function with reasoning
-def generate_text(input_text):
-    if load_error or tokenizer is None or model is None:
-        return f"Model failed to load: {load_error}", "Unable to proceed due to model loading error."
+with gr.Blocks(fill_height=True) as demo:
+    with gr.Sidebar():
+        gr.Markdown("# Inference Provider")
+        gr.Markdown("This Space showcases the agentica-org/DeepScaleR-1.5B-Preview model, served by the hf-inference API. Sign in with your Hugging Face account to use this API.")
+        button = gr.LoginButton("Sign in")
+    gr.load("models/agentica-org/DeepScaleR-1.5B-Preview", accept_token=button, provider="hf-inference")
     
-    try:
-        reasoning = "Step 1: Tokenizing the input text...\n"
-        inputs = tokenizer(input_text, return_tensors="pt").to(device)
-        reasoning += f"Input tokenized into {inputs['input_ids'].shape[1]} tokens.\n"
-
-        reasoning += "Step 2: Running the DeepscaleR model for generation...\n"
-        outputs = model.generate(
-            inputs["input_ids"],
-            max_length=100,
-            num_return_sequences=1,
-            temperature=0.7,
-            do_sample=True
-        )
-        reasoning += f"Generated {outputs.shape[1]} tokens.\n"
-
-        reasoning += "Step 3: Decoding the output tokens into readable text...\n"
-        generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
-        reasoning += "Decoding complete.\n"
-
-        reasoning += "Step 4: Finalizing the response.\n"
-        return generated_text, reasoning
-    except Exception as e:
-        error_msg = f"Error: {str(e)}"
-        return error_msg, f"Failed due to: {error_msg}"
-
-# Custom CSS for black theme
-custom_css = """
-body { background-color: #1a1a1a; color: #ffffff; }
-.gr-box { background-color: #2b2b2b; border: 1px solid #444444; border-radius: 5px; }
-.gr-button { background-color: #4a4a4a; color: #ffffff; border: none; }
-.gr-button:hover { background-color: #5a5a5a; }
-.gr-textbox, .gr-textarea { background-color: #333333; color: #ffffff; border: 1px solid #555555; }
-h1, h2, h3 { color: #ffffff; }
-"""
-
-# Gradio interface
-with gr.Blocks(css=custom_css, theme="default") as demo:
-    gr.Markdown(
-        """
-        # DeepscaleR Model Demo
-        A sleek, professional interface powered by xAI's Grok-inspired design.
-        Enter text below to see the DeepscaleR model's output and reasoning process.
-        """
-    )
-    
-    with gr.Row():
-        with gr.Column(scale=1):
-            input_text = gr.Textbox(
-                label="Input Text",
-                placeholder="Type your input here...",
-                lines=3
-            )
-            submit_btn = gr.Button("Generate")
-        
-        with gr.Column(scale=2):
-            output_text = gr.Textbox(label="Generated Output", lines=5)
-            reasoning_text = gr.Textbox(label="Reasoning Process", lines=10)
-
-    submit_btn.click(
-        fn=generate_text,
-        inputs=input_text,
-        outputs=[output_text, reasoning_text]
-    )
-
-# Launch the app
 demo.launch()