Enable NUMA feature for llama_cpp_python (#4040)
This commit is contained in:
parent
87ea2d96fd
commit
7e6ff8d1f0
7 changed files with 21 additions and 12 deletions
|
@ -82,7 +82,8 @@ def list_model_elements():
|
|||
'max_seq_len',
|
||||
'compress_pos_emb',
|
||||
'alpha_value',
|
||||
'rope_freq_base'
|
||||
'rope_freq_base',
|
||||
'numa',
|
||||
]
|
||||
|
||||
for i in range(torch.cuda.device_count()):
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue