Experimental 11B instruction tuned model based on flan-t5 Reference Finetuned Language Models Are Zero-Shot Learners