Merge branch 'main' into attention
29038ea
duzx16
commited on
Commit History
Add support for flash attention 2
a7eaddd
duzx16
commited on
Add output_attentions for ChatGLMForSequenceClassification
f07eca1
duzx16
commited on
Add eager and sdpa attention implementations
835c717
duzx16
commited on
Merge branch 'main' of https://huggingface.co/THUDM/glm-4-9b-chat
dba7772
duzx16
commited on
Fix default dtype for classification head
0deb1dd
duzx16
commited on
disable `torch.jit` when using Ascend NPUs (#20)
7d93f12
verified
Fix classification model
12c8049
duzx16
commited on
8dc9e30
verified
fixes https://github.com/THUDM/GLM-4/issues/22 (#5)
0891486
verified
first commit
4b556ad
Ubuntu
commited on