2 ms·
a new T5 * llama tokenizer * the pile dataset * trained on 2T tokens (2 times of og T5) * better, & especially better at coding
by tosh 3y ago
a new T5
* llama tokenizer
* the pile dataset
* trained on 2T tokens (2 times of og T5)
* better, & especially better at coding