Update app.py

app.py

@@ -1,4 +1,4 @@
-# Multi-agent UI generator with Hugging Face token-based gated model support
+# Multi-agent UI generator with token handling and model load debug support
 
 import streamlit as st
 import time
@@ -33,7 +33,7 @@ AGENT_MODEL_CONFIG = {
         "adapter": "spandana30/software-engineer-codellama"
     },
     "qa": {
-        "base": "codellama/CodeLlama-7b-Instruct-hf",
+        "base": "codellama/CodeLLaMA-7b-Instruct-hf",
         "adapter": "spandana30/software-engineer-codellama"
     },
 }
@@ -41,12 +41,21 @@ AGENT_MODEL_CONFIG = {
 @st.cache_resource
 
 def load_agent_model(base_id, adapter_id):
-    base_model = AutoModelForCausalLM.from_pretrained(
-        base_id, torch_dtype=torch.float16, device_map="auto", token=HF_TOKEN
-    )
-    model = PeftModel.from_pretrained(base_model, adapter_id, token=HF_TOKEN)
-    tokenizer = AutoTokenizer.from_pretrained(adapter_id, token=HF_TOKEN)
-    return pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=1024)
+    try:
+        st.write(f"🔍 Loading: {base_id} with adapter {adapter_id}")
+        st.write(f"🔐 Using token: {'Yes' if HF_TOKEN else 'No'}")
+
+        base_model = AutoModelForCausalLM.from_pretrained(
+            base_id, torch_dtype=torch.float16, device_map="auto", token=HF_TOKEN
+        )
+        model = PeftModel.from_pretrained(base_model, adapter_id, token=HF_TOKEN)
+        tokenizer = AutoTokenizer.from_pretrained(adapter_id, token=HF_TOKEN)
+
+        return pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=1024)
+
+    except Exception as e:
+        st.error(f"❌ Model load failed!\nBase: {base_id}\nAdapter: {adapter_id}\nError: {e}")
+        raise
 
 AGENT_PIPELINES = {
     role: load_agent_model(cfg["base"], cfg["adapter"])