192 6 days ago

Latest RWKV-7 reasoning models.

thinking 1.5b 2.9b 7.2b 13.3b
b97d2f6c3103 · 124B
{
"frequency_penalty": 0.5,
"presence_penalty": 0.5,
"repeat_penalty": 1,
"stop": [
"\n\n"
],
"temperature": 1,
"top_k": 128,
"top_p": 0.5
}