SEBIS/code_trans_t5_small_source_code_summarization_csharp_multitask_finetune

The CodeTrans model is based on the t5-small model architecture . It has its own SentencePiece vocabulary model . It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets . It is then fine-tuned on the source code summarization task for the csharp code snippets . The model could be used to generate the description for a csharp function . It works best with tokenized csharp functions . It can be used on unparsed and untokenized code, but the performance should be better if the code is tokenized . The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training,