from sentence_transformers import SentenceTransformer import torch print("🚀 Warming up model...") embedding_model = SentenceTransformer("/app/model_cache", device="cpu") embedding_model = embedding_model.half() # Reduce memory embedding_model.to(torch.device("cpu")) print("✅ Model warm-up complete!")