Using useage is an error encountered

[JL-er]

from gla import GatedLinearAttention import torch d_model = 1024 num_head = 4 use_gk = True # alpha use_gv = False # beta device = "cuda:0"

gla_layer = GatedLinearAttention(d_model, num_head, use_gk, use_gv).to(device)

bsz, seq_len, d_model = 32, 2048, 1024 x = torch.randn(bsz, seq_len, d_model).to(device) y = gla_layer(x)

assert y.shape == x.shape

[sustcsonglin]

Hello, which triton version are you using? Can you delete the triton cache file? which should be located in ~/.triton

[JL-er]

Hello, which triton version are you using? Can you delete the triton cache file? which should be located in ~/.triton

Thanks, the problem was solved after I updated the version