DeepSeek's first-generation of reasoning models with comparable performance to OpenAI-o1, including six dense models distilled from DeepSeek-R1 based on Llama and Qwen.

1.5b 7b 8b 14b 32b 70b 671b

8.7M 2 weeks ago

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