Deploy LaunchLLM - Production AI Training Platform

README.md

@@ -4,7 +4,7 @@ emoji: 🚀
 colorFrom: blue
 colorTo: purple
 sdk: gradio
-sdk_version: 5.49.1
+sdk_version: 4.0.0
 app_file: app.py
 pinned: false
 license: apache-2.0
@@ -174,4 +174,4 @@ Start by clicking the **Environment** tab above and adding your HuggingFace toke
 
 ---
 
-**Built with ❤️ for domain experts who want custom AI without the complexity**
+**Built with ❤️ for domain experts who want custom AI without the complexity**

app.py

@@ -3,8 +3,9 @@ Hugging Face Spaces deployment entry point for AURA AI Training Lab
 This file is automatically used by HF Spaces to launch the application.
 """
 
-# Import and launch the main GUI
+# Import the GUI module to get the demo object
 import financial_advisor_gui
 
-# The demo.launch() call at the bottom of financial_advisor_gui.py
-# will automatically run when this module is imported
+# Get the demo object and launch it
+# HuggingFace Spaces will automatically detect this
+demo = financial_advisor_gui.demo

financial_advisor_gui.py

@@ -1897,7 +1897,7 @@ Search: "finance", "financial", "investment", "trading\"""",
                     model_choices = model_registry.get_model_choices_for_gui()
                     model_selector = gr.Dropdown(
                         choices=model_choices,
-                        value=model_choices[0],  # Default to first model (Qwen 3 30B)
+                        value=model_choices[0][1] if model_choices else None,  # Default to first model ID
                         label="Select Model",
                         info="Choose which model to train"
                     )
@@ -1987,7 +1987,7 @@ Search: "finance", "financial", "investment", "trading\"""",
             # Wire up training
             start_train_btn.click(
                 fn=start_training,
-                inputs=[lora_rank, learning_rate, num_epochs, batch_size, grad_accum],
+                inputs=[lora_rank, learning_rate, num_epochs, batch_size, grad_accum, training_mode],
                 outputs=[training_log, training_status_text]
             )