implementation detail · filed under data curation
Separate PT and IT end tokens
A formatting distinction in which pre-training models end with <eos> and instruction-tuned models end with <turn|>.
Also called PT versus IT formatting.
- source
- 1
- model
- 1
- lab adopt it
- 1
- strongest
- used
How sources treat it
One count per evidence span, weakest treatment to strongest.
used 1
Documented in
Evidence
1 span quoted from the sources, strongest treatment first.
PT models output an <eos> token at the end of generation, while IT models output <turn|> at the end of the generation.
usedpost trainingin Gemma 4Google DeepMind
Filed alongside
Other methods under data curation :: tokenization.