Context window increase to 768

app.py

@@ -15,7 +15,7 @@ headers = {
 }
 
 hf_hub_download(repo_id="LLukas22/gpt4all-lora-quantized-ggjt", filename="ggjt-model.bin", local_dir=".")
-llm = Llama(model_path="./ggjt-model.bin")
+llm = Llama(model_path="./ggjt-model.bin", n_ctx=768)
 
 ins = '''### Instruction:
 {question}