Additional training on Japanese data by CyberAgent for deepseek-r1.

14b 32b

2,191 3 months ago

f4d24e9138dd · 148B
{
"stop": [
"<|begin▁of▁sentence|>",
"<|end▁of▁sentence|>",
"<|User|>",
"<|Assistant|>"
]
}