Update app.py

app.py

@@ -25,7 +25,8 @@ MODELS = [
     'MihaiPopa-1/CinnabarLM-1.5M-Base', 'Harley-ml/Dillionv2-1.3M', 'Eclipse-Senpai/KeyLM-75M',
     'SupraLabs/Supra-Mini-v6-1M', 'AxiomicLabs/GPT-S-1.4M', 'GODELEV/Archaea-74M',
     'Sandroeth/cali-0.1B', 'veyra-ai/veyra3-5m-base', 'veyra-ai/veyra-30m-base-5b-tokens',
-    'ThingAI/Quark-50m', 'ThingAI/Quark-135m', 'HuggingFaceTB/SmolLM2-135M-Instruct'
+    'ThingAI/Quark-50m', 'ThingAI/Quark-135m', 'HuggingFaceTB/SmolLM2-135M-Instruct', 
+    'Aravindan/awesome-gpt-2-coder', 'Qwen/Qwen2.5-Coder-0.5B'
 ]
 
 # Global class to safely manage the loaded model and tokenizer in memory
@@ -52,6 +53,7 @@ def load_new_model(model_id):
     # Clear old model from memory
     model_manager.model = None
     model_manager.tokenizer = None
+    yield f"Loading {model_id}..."
     gc.collect()
     if torch.cuda.is_available():
         torch.cuda.empty_cache()
@@ -64,9 +66,9 @@ def load_new_model(model_id):
         model_manager.tokenizer = tokenizer
         model_manager.model = model
         
-        return f"Successfully loaded {model_id} on {model_manager.device.upper()}"
+        yield f"Successfully loaded {model_id} on {model_manager.device.upper()}"
     except Exception as e:
-        return f"Error loading model: {str(e)}"
+        yield f"Error loading model: {str(e)}"
 
 def run_inference(user_prompt, max_tokens, temperature, top_k, top_p, rep_penalty, ngram_size, do_sample):
     """Generates text via streaming generator."""