Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
gaoqiong
lm-evaluation-harness
Commits
bbd6d051
"mmdet3d/vscode:/vscode.git/clone" did not exist on "809d1a9b009cb4333f9414427a079854f7f50852"
Commit
bbd6d051
authored
Nov 27, 2023
by
baberabb
Browse files
add auto batching
parent
37f10cad
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
0 additions
and
5 deletions
+0
-5
lm_eval/models/vllm_causallms.py
lm_eval/models/vllm_causallms.py
+0
-5
No files found.
lm_eval/models/vllm_causallms.py
View file @
bbd6d051
...
@@ -36,11 +36,6 @@ class VLLM(LM):
...
@@ -36,11 +36,6 @@ class VLLM(LM):
):
):
super
().
__init__
()
super
().
__init__
()
assert
"cuda"
in
device
or
device
is
None
,
"vLLM only supports CUDA"
assert
"cuda"
in
device
or
device
is
None
,
"vLLM only supports CUDA"
if
batch_size
==
"auto"
:
eval_logger
.
info
(
"vllm does not support auto selection. Setting batch_size to 8"
)
batch_size
=
8
self
.
model
=
LLM
(
self
.
model
=
LLM
(
model
=
pretrained
,
model
=
pretrained
,
gpu_memory_utilization
=
gpu_memory_utilization
,
gpu_memory_utilization
=
gpu_memory_utilization
,
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment