update

app.py

@@ -188,12 +188,37 @@ def get_llm(cfg):
 
 
 def run(text, intensity):
-    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)
+        # 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)
 
-    _ = language_model.generate(
+        sample_outputs = language_model.generate(
             **input_ids,
             streamer=streamer,
             pad_token_id=tokenizer.pad_token_id,
@@ -202,19 +227,33 @@ def run(text, intensity):
             temperature=cfg.generation.temperature
         )
 
-    # print(20*'---')
+        
+        # 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)
 
 
-    res="Chatbot Data Mining 2024   \n  \n  \n"
-    max_length=intensity
 
-    return _
+
+    # print(20*'---')
+
+    # res="Chatbot Data Mining 2024   \n  \n  \n"
+    # max_length=intensity
+
+    return sample_outputs
 
 
-def vistral_chat(cfg, retriever, tokenizer, language_model):
+def vistral_chat():
     demo = gr.Interface(fn=run,
-                    inputs=[gr.Textbox(label="Nhập vào nội dung input",value="Con đường xưa em đi"),gr.Slider(label="Độ dài output muốn tạo ra", value=20, minimum=10, maximum=100, step=2)],
+                    inputs=[gr.Textbox(label="Nhập vào nội dung input",value="Con đường xưa em đi"),gr.Slider(label="Độ dài output muốn tạo ra", value=20, minimum=10, maximum=100, step=2),],
                     outputs=gr.Textbox(label="Output"),  # <-- Number of output components: 1
                     )