Update app.py

app.py

@@ -109,10 +109,20 @@ class LLMResponder:
             self.llm = Llama(model_path="path/to/your/gguf/file.gguf", n_ctx=1024)
             self.backend = "llama_cpp"
         else:
-            # Create a dummy config using LlamaConfig so that the model loads even if its config is missing a model_type key.
+            # Create a dummy config using LlamaConfig so the model loads despite missing keys.
             dummy_config = LlamaConfig.from_dict({"model_type": "llama"})
-            self.llm_tokenizer = AutoTokenizer.from_pretrained(model_name, config=dummy_config, trust_remote_code=True)
-            self.llm_model = AutoModelForCausalLM.from_pretrained(model_name, config=dummy_config, trust_remote_code=True)
+            try:
+                self.llm_tokenizer = AutoTokenizer.from_pretrained(model_name, config=dummy_config, trust_remote_code=True)
+            except Exception as e:
+                print("Error loading tokenizer from", model_name, "; using fallback tokenizer.")
+                fallback_model = "decapoda-research/llama-7b-hf"
+                self.llm_tokenizer = AutoTokenizer.from_pretrained(fallback_model, config=dummy_config, trust_remote_code=True)
+            try:
+                self.llm_model = AutoModelForCausalLM.from_pretrained(model_name, config=dummy_config, trust_remote_code=True)
+            except Exception as e:
+                print("Error loading model from", model_name, "; using fallback model.")
+                fallback_model = "decapoda-research/llama-7b-hf"
+                self.llm_model = AutoModelForCausalLM.from_pretrained(fallback_model, config=dummy_config, trust_remote_code=True)
             self.backend = "transformers"
 
     def generate_response(self, prompt):