Closed epsilon-deltta closed 2 years ago
Hi @epsilon-deltta, at current stage, SummerTime does not support fine-tuning/pre-training yet. Though we see this as important future work, it is unlikely that we are going to complete it in the next month or two.
Please refer to other implementations, such as huggingface or fairseq.
Closed due to inactivity.
I want to do BART finetuning on my dataset, Could you provide a snippet of the train code? And, I also want to do pre-training on another dataset. If there is, could you send me a link.