chche dir

app/rag_pipeline/model_initializer.py

@@ -32,12 +32,12 @@ def initialize_models(openai_api_key=None,model_id=None, model_basename=None):
                                                     encode_kwargs={'normalize_embeddings': False},
                                                     cache_folder = CACHE_DIR
                                                     )
-            # llm_model = load_model(device_type="cpu", model_id=model_id, model_basename=model_basename, LOGGING=logger)
-            llm_model = HuggingFacePipeline.from_model_id(
-                model_id= "gpt2", #"google/flan-t5-small",
-                task="text-generation",
-                cache_dir=CACHE_DIR,
-            )
+            llm_model = load_model(device_type="cpu", model_id=model_id, model_basename=model_basename, LOGGING=logger)
+            # llm_model = HuggingFacePipeline.from_model_id(
+            #     model_id= "gpt2", #"google/flan-t5-small",
+            #     task="text-generation",
+            #     cache_dir=CACHE_DIR,
+            # )
 
             #TheBloke/Mistral-7B-v0.1-GGUF
             #HuggingFaceH4/zephyr-7b-beta