Feature Enhancement: Batch Inference Support in candle-binding (#71)

* Feature Enhancement: Batch Inference Support in candle-binding

Feature Enhancement: Batch Inference Support in candle-binding

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

fix: unified_classifier_test

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

fix: unified_classifier_test

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

fix: unit_test

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

* 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]>

* fix: unit test and model download from huggingface

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

fix: unit test and model download from huggingface

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

fix: unit test and model download from huggingface

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

fix: unit test and model download from huggingface

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

fix: unit test and model download from huggingface

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

fix: unit test and model download from huggingface

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

fix: unit test and model download from huggingface

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

---------

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

Author: OneZero-Y

Makefile

@@ -256,6 +256,42 @@ 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
 
+	@if [ ! -d "lora_intent_classifier_bert-base-uncased_model" ]; then \
+		hf download LLM-Semantic-Router/lora_intent_classifier_bert-base-uncased_model --local-dir models/lora_intent_classifier_bert-base-uncased_model; \
+	fi
+
+	@if [ ! -d "models/lora_intent_classifier_roberta-base_model" ]; then \
+		hf download LLM-Semantic-Router/lora_intent_classifier_roberta-base_model --local-dir models/lora_intent_classifier_roberta-base_model; \
+	fi
+
+	@if [ ! -d "models/lora_intent_classifier_modernbert-base_model" ]; then \
+		hf download LLM-Semantic-Router/lora_intent_classifier_modernbert-base_model --local-dir models/lora_intent_classifier_modernbert-base_model; \
+	fi
+
+	@if [ ! -d "models/lora_pii_detector_bert-base-uncased_model" ]; then \
+		hf download LLM-Semantic-Router/lora_pii_detector_bert-base-uncased_model --local-dir models/lora_pii_detector_bert-base-uncased_model; \
+	fi
+
+	@if [ ! -d "models/lora_pii_detector_roberta-base_model" ]; then \
+		hf download LLM-Semantic-Router/lora_pii_detector_roberta-base_model --local-dir models/lora_pii_detector_roberta-base_model; \
+	fi
+
+	@if [ ! -d "models/lora_pii_detector_modernbert-base_model" ]; then \
+		hf download LLM-Semantic-Router/lora_pii_detector_modernbert-base_model --local-dir models/lora_pii_detector_modernbert-base_model; \
+	fi
+
+	@if [ ! -d "models/lora_jailbreak_classifier_bert-base-uncased_model" ]; then \
+		hf download LLM-Semantic-Router/lora_jailbreak_classifier_bert-base-uncased_model --local-dir models/lora_jailbreak_classifier_bert-base-uncased_model; \
+	fi
+
+	@if [ ! -d "models/lora_jailbreak_classifier_roberta-base_model" ]; then \
+		hf download LLM-Semantic-Router/lora_jailbreak_classifier_roberta-base_model --local-dir models/lora_jailbreak_classifier_roberta-base_model; \
+	fi
+
+	@if [ ! -d "models/lora_jailbreak_classifier_modernbert-base_model" ]; then \
+		hf download LLM-Semantic-Router/lora_jailbreak_classifier_modernbert-base_model --local-dir models/lora_jailbreak_classifier_modernbert-base_model; \
+	fi
+
 # Milvus container management
 start-milvus:
 	@echo "Starting Milvus container for testing with $(CONTAINER_RUNTIME)..."