tweak 2

app.py

@@ -7,23 +7,13 @@ from huggingface_hub import hf_hub_download
 # huggingface-cli download microsoft/Phi-3-mini-4k-instruct-gguf Phi-3-mini-4k-instruct-q4.gguf --local-dir .
 # huggingface-cli download LoneStriker/OpenBioLLM-Llama3-8B-GGUF --local-dir ./llama3-gguf
 
-# Define the model name and file
-model_name = "LoneStriker/OpenBioLLM-Llama3-8B-GGUF"
-model_file = "Llama3-8B-Q5_K_M.gguf"
-
-# Define the local directory path within the Docker container
-local_dir = "/usr/src/app/llama3-gguf"
-
-# Ensure that the local directory exists
-os.makedirs(local_dir, exist_ok=True)
-
-# Download the model to the specified local directory
-model_path = hf_hub_download(model_name, filename=model_file, local_dir=local_dir)
-
 llm = Llama(
     # model_path="./Phi-3-mini-4k-instruct-q4.gguf",
     # model_path="./llama3-gguf/OpenBioLLM-Llama3-8B-Q5_K_M.gguf",
-    model_path = model_path,
+    model_path = hf_hub_download(
+        repo_id=os.environ.get("REPO_ID", "LoneStriker/OpenBioLLM-Llama3-8B-GGUF"),
+        filename=os.environ.get("MODEL_FILE", "Llama3-8B-Q5_K_M.gguf"),
+    ),
     n_ctx=2048,
     n_gpu_layers=50, # change n_gpu_layers if you have more or less VRAM 
 )