Update app.py

app.py

@@ -6,8 +6,8 @@ from llama_cpp import Llama
 from huggingface_hub import hf_hub_download #load from huggingfaces 
 
 
-llm = Llama(model_path= hf_hub_download(repo_id="TheBloke/Vigogne-2-7B-Instruct-GGML", filename="vigogne-2-7b-instruct.ggmlv3.q4_1.bin"))
-llm.load_model() #pre load model(as suggested by chatGPT)
+llm = Llama(model_path= hf_hub_download(repo_id="TheBloke/Vigogne-2-7B-Instruct-GGML", filename="vigogne-2-7b-instruct.ggmlv3.q4_1.bin"), n_ctx=2048) #download model from hf/ n_ctx=2048 for high ccontext length
+
 
 def generate_text(input_text):
     output = llm(f"Q: {input_text}  A:", max_tokens=521, stop=["Q:", "\n"], echo=True)