Update app.py

app.py

@@ -3,10 +3,10 @@ import gradio as gr
 from llama_cpp import Llama
 from huggingface_hub import snapshot_download
 
-# Optionally, set a high timeout environment variable (if your HF Hub version honors this)
-os.environ["HF_HUB_HTTP_TIMEOUT"] = "3600"  # 1 hour in seconds
+# Optionally increase the HTTP timeout via an environment variable
+os.environ["HF_HUB_HTTP_TIMEOUT"] = "3600"  # 1 hour
 
-# Download the model repository locally; this will block until the download completes
+# Download the model repository locally; this call blocks until the download completes.
 local_repo = snapshot_download(
     repo_id="DebabrataHalder/mysqlmodel",
     revision="main",
@@ -14,20 +14,19 @@ local_repo = snapshot_download(
     resume_download=True
 )
 
-# Load the model from the local repository
+# Load the model using the remote repo_id and specify the local cache directory
 llm = Llama.from_pretrained(
-    repo_id=local_repo,           # local folder path now
-    filename="mysqlmodel.gguf",   # the GGUF file name
-    n_ctx=2048,                   # adjust context size if needed
-    local_files_only=True         # force local file loading
+    repo_id="DebabrataHalder/mysqlmodel",  # valid repo id format
+    filename="mysqlmodel.gguf",
+    cache_dir=local_repo,                   # point to the downloaded snapshot
+    n_ctx=2048,                           # adjust context size if needed
+    local_files_only=True                   # ensure only local files are used
 )
 
 def chat(prompt):
-    # Generate a response using the model
     output = llm(prompt)
     return output["choices"][0]["text"]
 
-# Create Gradio interface
 iface = gr.Interface(
     fn=chat,
     inputs="text",
@@ -38,4 +37,3 @@ iface = gr.Interface(
 
 iface.launch()
 
-