Training data format

[StarGazerM]

Hi :
I want to train stateformer with my own dateset, can you provide some detail explaination of data format in data-src? what format is *.byte* and what is .inst_pos_emb, how can I extract them from DWARF ?

[analognahid]

I understand what is the inst_pos_emb. For example '0 0 0 0 0 0 0 1 1 1 1 1 1 1 2 2 2 2 2 2 2 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 ' means the 1st 6 tokens belong to 0th instruction, the next 6 to 1st and so on. I still don't understand the .byye format.