A variant of the DeepSeek-R1 model, post-trained by Perplexity to deliver unbiased, accurate, and factual information.
117 Pulls Updated 3 weeks ago
edbb4306f7ca · 132B
{
"stop": [
"<|begin▁of▁sentence|>",
"<|end▁of▁sentence|>",
"<|User|>",
"<|Assistant|>"
]
}