update

app.py

@@ -188,37 +188,13 @@ def get_llm(cfg):
 
 
 def run(text, intensity):
-        # Configure logging
-    logging.basicConfig(level=logging.INFO, 
-                        format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
-    logger = logging.getLogger(__name__)
-    sample_outputs = "start: "
-    
-    try:
         # Log the start of the process
-        logger.info("Starting the process with config file: %s", config_path)
-        
-        # Load configuration from the file
-        config = load_config(config_path)
-        
-        # Load necessary components
-        prompt_template = get_prompt_template()
-        
-        # Replace OpenAI embed model and llm with custom ones
-        reset_settings(config)
-        
-        # Get retriever
-        retriever = get_retriever(config, prompt_template)
-        
-        # Load tokenizer and language model
-        tokenizer = load_tokenizer(config)
-        language_model = get_llm(config)
-        prompt = retriever.query(text).response
-        prompt = tokenizer.bos_token + '[INST] ' + prompt + ' [/INST]'
-        streamer = TextStreamer(tokenizer, skip_prompt=True)
-        input_ids = tokenizer([prompt], return_tensors='pt').to(cfg.environment.device)
+    prompt = retriever.query(text).response
+    prompt = tokenizer.bos_token + '[INST] ' + prompt + ' [/INST]'
+    streamer = TextStreamer(tokenizer, skip_prompt=True)
+    input_ids = tokenizer([prompt], return_tensors='pt').to(cfg.environment.device)
 
-        sample_outputs = language_model.generate(
+    sample_outputs = language_model.generate(
             **input_ids,
             streamer=streamer,
             pad_token_id=tokenizer.pad_token_id,
@@ -227,27 +203,6 @@ def run(text, intensity):
             temperature=cfg.generation.temperature
         )
 
-        
-        # Start the command line interface
-        # vistral_chat(config, retriever, tokenizer, language_model)
-        
-        # Log successful completion
-        logger.info("Process completed successfully.")
-        
-    except FileNotFoundError as e:
-        logger.error("Configuration file not found: %s", e)
-
-    except Exception as e:
-        logger.exception("An error occurred: %s", e)
-
-
-
-
-    # print(20*'---')
-
-    # res="Chatbot Data Mining 2024   \n  \n  \n"
-    # max_length=intensity
-
     return sample_outputs
 
 
@@ -261,23 +216,6 @@ def vistral_chat():
 
 
 
-    # while True:
-    #     user_query = input('👨‍🦰 ')
-    #     prompt = retriever.query(user_query).response
-    #     prompt = tokenizer.bos_token + '[INST] ' + prompt + ' [/INST]'
-    #     streamer = TextStreamer(tokenizer, skip_prompt=True)
-    #     input_ids = tokenizer([prompt], return_tensors='pt').to(cfg.environment.device)
-
-    #     _ = language_model.generate(
-    #         **input_ids,
-    #         streamer=streamer,
-    #         pad_token_id=tokenizer.pad_token_id,
-    #         max_new_tokens=cfg.generation.max_new_tokens,
-    #         do_sample=cfg.generation.do_sample,
-    #         temperature=cfg.generation.temperature
-    #     )
-
-    #     print(20*'---')
 
 
 def main1(config_path):
@@ -285,7 +223,10 @@ def main1(config_path):
     logging.basicConfig(level=logging.INFO, 
                         format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
     logger = logging.getLogger(__name__)
-    
+    global config = None 
+    global retriever=None
+    global tokenizer=None
+    global language_model= None
     try:
         # Log the start of the process
         logger.info("Starting the process with config file: %s", config_path)
@@ -307,7 +248,7 @@ def main1(config_path):
         language_model = get_llm(config)
         
         # Start the command line interface
-        vistral_chat(config, retriever, tokenizer, language_model)
+        vistral_chat()
         
         # Log successful completion
         logger.info("Process completed successfully.")