SEBIS/code_trans_t5_base_program_synthese_multitask

The CodeTrans model is based on the t5-base model architecture . It has its own SentencePiece vocabulary model . It could be used to generate lisp inspired DSL code given the human language description tasks . The model was trained on a single TPU Pod V3-8 for 360,000 steps in total, using sequence length 512 (batch size 512) The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training . The training data can be downloaded on Link: Link . The code documentation tasks are evaluated using different models on different programming languages (in BLEU score) The results are based on various programming languages, with different models achieving the,