Synced repo using 'sync_with_huggingface' Github Action

Dockerfile

@@ -82,8 +82,8 @@ EXPOSE 7860
 
 # Start the FastAPI app using Uvicorn web server
 # CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "14000", "--limit-concurrency", "1000"]
-RUN python3 download.py
+# RUN python3 download.py
 
 # RUN chmod 755 models
 
-CMD ["python3", "app.py", "--host=0.0.0.0", "--port=7860", "--model_path=models/BAAI/bge-small-en-v1.5", "--num_workers=2"]
+CMD ["python3", "app.py", "--host=0.0.0.0", "--port=7860", "--model_path=BAAI/bge-small-en-v1.5", "--num_workers=2"]

app.py

@@ -70,9 +70,9 @@ print_arguments(args)
 assert os.path.exists(args.model_path), f"{args.model_path}"
 # 
 if args.use_gpu:
-    model = SentenceTransformer(args.model_path, device="cuda", compute_type="float16")
+    model = SentenceTransformer(args.model_path, device="cuda", compute_type="float16", cache_folder=".")
 else:
-    model = SentenceTransformer(args.model_path, device='cpu')
+    model = SentenceTransformer(args.model_path, device='cpu', cache_folder=".")
 
 
 app = FastAPI(title="embedding Inference")