Update app.py

app.py

@@ -10,24 +10,26 @@ import subprocess
 from huggingface_hub import hf_hub_download
 from llama_cpp import Llama
 
+model_name = "Meta-Llama-3-70B-Instruct-Q3_K_M.gguf"
+
 # hf_hub_download(
 #     repo_id="QuantFactory/Meta-Llama-3-8B-Instruct-GGUF",
 #     filename="Meta-Llama-3-8B-Instruct.Q8_0.gguf",
 #     local_dir = "./models"
 # )
 
-# hf_hub_download(
-#     repo_id="bartowski/Meta-Llama-3-70B-Instruct-GGUF",
-#     filename="Meta-Llama-3-70B-Instruct-Q3_K_M.gguf",
-#     local_dir = "./models"
-# )
-
 hf_hub_download(
-    repo_id="leafspark/Meta-Llama-3.1-405B-Instruct-GGUF",
-    filename="Llama-3.1-405B-Instruct.Q2_K.gguf",
+    repo_id="bartowski/Meta-Llama-3-70B-Instruct-GGUF",
+    filename=model_name,
     local_dir = "./models"
 )
 
+# hf_hub_download(
+#     repo_id="leafspark/Meta-Llama-3.1-405B-Instruct-GGUF",
+#     filename="Llama-3.1-405B-Instruct.Q2_K.gguf",
+#     local_dir = "./models"
+# )
+
 def process_document(pdf_path):
     extracted_pages = extract_pages(pdf_path)
     page2content = {}
@@ -99,7 +101,7 @@ def pdf_to_text(cv_file, prompt, maxtokens=2048, temperature=0, top_probability=
     for page_id in page2content:
         cv_text += page2content[page_id] + ' '
     llm = Llama(
-        model_path="models/Llama-3.1-405B-Instruct.Q2_K.gguf",
+        model_path="models/" + model_name,
         flash_attn=True,
         n_gpu_layers=81,
         n_batch=1024,