feat: add LoRA training pipeline and intelligent model selection

- Complete LoRA training scripts for 3 classification tasks
- Smart model selection with architecture priority (BERT > RoBERTa > ModernBERT)
- Official Candle BERT integration for Python-Go consistency
- Enhanced unified classifier with high-confidence LoRA models

Signed-off-by: OneZero-Y <[email protected]>

Author: OneZero-Y

Makefile

@@ -256,6 +256,13 @@ download-models:
 		hf download LLM-Semantic-Router/pii_classifier_modernbert-base_presidio_token_model --local-dir models/pii_classifier_modernbert-base_presidio_token_model; \
 	fi
 
+	# LoRA Enhanced Models (Note: These need to be trained locally)
+	@echo "LoRA models need to be trained locally using the training scripts:"
+	@echo "  - Intent Classification: python src/training/training_lora/classifier_model_fine_tuning_lora/ft_linear_lora.py"
+	@echo "  - PII Detection: python src/training/training_lora/pii_model_fine_tuning_lora/pii_bert_finetuning_lora.py"
+	@echo "  - Security Detection: python src/training/training_lora/prompt_guard_fine_tuning_lora/jailbreak_bert_finetuning_lora.py"
+	@echo "See website/docs/training/training-overview.md for detailed instructions."
+
 # Milvus container management
 start-milvus:
 	@echo "Starting Milvus container for testing with $(CONTAINER_RUNTIME)..."