Hello,
I was wondering if you could release the code for the T5 large 770M fine-tuned SLM with the dataset that allow to replicate the performance of the paper? Have you tried to fine-tune T5-XXL or even Flan-T5 for an even greater perf boost on the same dataset on which the T5 large has been trained?
Hello,
I was wondering if you could release the code for the T5 large 770M fine-tuned SLM with the dataset that allow to replicate the performance of the paper? Have you tried to fine-tune T5-XXL or even Flan-T5 for an even greater perf boost on the same dataset on which the T5 large has been trained?