Update app.py

app.py

@@ -11,6 +11,7 @@ from gradio_huggingfacehub_search import HuggingfaceHubSearch
 from apscheduler.schedulers.background import BackgroundScheduler
 from datetime import datetime
 import numpy as np
+import shutil
 HF_TOKEN = os.environ.get("HF_TOKEN")
 
 os.environ["GRADIO_ANALYTICS_ENABLED"] = "False"
@@ -204,7 +205,7 @@ def process_model(model_id, q_method, use_imatrix, imatrix_q_method, private_rep
 
     current_time = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
     logger.info(f"Time {current_time}, Username {username}, Model_ID, {model_id}, q_method {','.join(q_method)}")
-
+    
     repo_namespace = get_repo_namespace(repo_owner, username, user_orgs)
     model_name = model_id.split('/')[-1]
     try:
@@ -225,7 +226,8 @@ def process_model(model_id, q_method, use_imatrix, imatrix_q_method, private_rep
           fp16 = str(Path(outdir)/f"{model_name}.fp16.gguf")
 
           with tempfile.TemporaryDirectory(dir=downloads_dir) as tmpdir:
-              print("Downloading")
+              print(datetime.now().strftime("%Y-%m-%d %H:%M:%S") + " Start download")
+              logger.info(datetime.now().strftime("%Y-%m-%d %H:%M:%S") + " Start download")
               local_dir = Path(tmpdir)/model_name
               api.snapshot_download(repo_id=model_id, local_dir=local_dir, local_dir_use_symlinks=False, allow_patterns=dl_pattern)
               
@@ -234,12 +236,15 @@ def process_model(model_id, q_method, use_imatrix, imatrix_q_method, private_rep
               if os.path.exists(adapter_config_dir) and not os.path.exists(config_dir):
                   raise Exception("adapter_config.json is present. If converting LoRA, use GGUF-my-lora.")
 
-              print("Download successfully")
+              print(datetime.now().strftime("%Y-%m-%d %H:%M:%S") + " Download successfully")
+              logger.info(datetime.now().strftime("%Y-%m-%d %H:%M:%S") + " Download successfully")
               result = subprocess.run(["python", CONVERSION_SCRIPT, local_dir, "--outtype", "f16", "--outfile", fp16], shell=False, capture_output=True)
               print("Converted to f16")
               if result.returncode != 0:
                   raise Exception(f"Error converting to fp16: {result.stderr.decode()}")
 
+          shutil.rmtree(downloads_dir)
+
           imatrix_path = Path(outdir)/"imatrix.dat"
           if use_imatrix:
               train_data_path = train_data_file.name if train_data_file else "llama.cpp/groups_merged.txt"
@@ -252,7 +257,8 @@ def process_model(model_id, q_method, use_imatrix, imatrix_q_method, private_rep
 
           gguf_files = []
           for method in quant_methods:
-              print("Begin quantize")
+              print(datetime.now().strftime("%Y-%m-%d %H:%M:%S") + " Begin quantize")
+              logger.info(datetime.now().strftime("%Y-%m-%d %H:%M:%S") + " Begin quantize")
               name = f"{model_name.lower()}-{method.lower()}-{suffix}.gguf" if suffix else f"{model_name.lower()}-{method.lower()}.gguf"
               path = str(Path(outdir)/name)
               quant_cmd = ["./llama.cpp/llama-quantize", "--imatrix", imatrix_path, fp16, path, method] if use_imatrix else ["./llama.cpp/llama-quantize", fp16, path, method]
@@ -262,7 +268,8 @@ def process_model(model_id, q_method, use_imatrix, imatrix_q_method, private_rep
               size = os.path.getsize(path)/1024/1024/1024
               gguf_files.append((name, path, size, method))
 
-          print("Quantize successfully!")
+          print(datetime.now().strftime("%Y-%m-%d %H:%M:%S") + " Quantize successfully!")
+          logger.info(datetime.now().strftime("%Y-%m-%d %H:%M:%S") + " Quantize successfully!")
           suffix_for_repo = f"{imatrix_q_method}-imat" if use_imatrix else "-".join(quant_methods)
           repo_id = f"{repo_namespace}/{model_name}-{suffix_for_repo}-GGUF"
           new_repo_url = api.create_repo(repo_id=repo_id, exist_ok=True, private=private_repo)