66RING
I don't know how to reproduce it. It seems crash randomly during runtime.  My environment: - OS: arch linux, 5.13.5 - CPU: AMD Ryzen 5 3550H with Radeon Vega...
just like this: ```python @triton.jit def kernel1(): .... def the_api(): data = kernel1() out = kernel2(data) ``` and I get `nan` when call `the_api` in a loop: ```python # test...
In my opinion, the generation should be the same when draft model and target model is the same and temparature is 0. But in this case, the output logits of...