Add --autosplit flag for ExLlamaV2 (#5524)

This commit is contained in:
oobabooga 2024-02-16 15:26:10 -03:00 committed by GitHub
parent 4039999be5
commit a6730f88f7
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
6 changed files with 29 additions and 17 deletions

View file

@ -76,6 +76,7 @@ def list_model_elements():
'no_flash_attn',
'num_experts_per_token',
'cache_8bit',
'autosplit',
'threads',
'threads_batch',
'n_batch',