Update app.py

app.py

@@ -113,15 +113,7 @@ db = Chroma(persist_directory=CHROMA_PATH, embedding_function=embeddings)
 
 
 
-generation_config = model.generation_config
-generation_config.temperature = 0
-generation_config.num_return_sequences = 1
-generation_config.max_new_tokens = 256
-generation_config.use_cache = False
-generation_config.repetition_penalty = 1.7
-generation_config.pad_token_id = tokenizer.eos_token_id
-generation_config.eos_token_id = tokenizer.eos_token_id
-generation_config
+
 
 
 prompt = """
@@ -160,16 +152,6 @@ stopping_criteria = StoppingCriteriaList(
     [StopGenerationCriteria(stop_tokens, tokenizer, model.device)]
 )
 
-generation_pipeline = pipeline(
-    model=model,
-    tokenizer=tokenizer,
-    return_full_text=True,
-    task="text-generation",
-    stopping_criteria=stopping_criteria,
-    generation_config=generation_config,
-)
-
-llm = HuggingFacePipeline(pipeline=generation_pipeline)
 
 
 class CleanupOutputParser(BaseOutputParser):