yingzhige00

about which attention used

2

which attention model the code use?

RuntimeError: Triton Error [CUDA]: context is destroyed

8

This is my program error: ```Mamba 1 took 0.7026619808748364 seconds Traceback (most recent call last): File "/root/mamba_test.py", line 34, in mamba2_time = timeit.timeit('try_mamba2(batch, length, dim, x)', number=10, globals=globals()) File "/opt/conda/envs/mamba2/lib/python3.10/timeit.py",...