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
OpenDAS
ollama
Commits
f32ea81b
Unverified
Commit
f32ea81b
authored
Jan 20, 2024
by
Jeffrey Morgan
Committed by
GitHub
Jan 20, 2024
Browse files
increase minimum overhead to 1024MiB (#2114)
parent
4c54f0dd
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
3 additions
and
3 deletions
+3
-3
gpu/gpu.go
gpu/gpu.go
+3
-3
No files found.
gpu/gpu.go
View file @
f32ea81b
...
...
@@ -191,11 +191,11 @@ func getCPUMem() (memInfo, error) {
func
CheckVRAM
()
(
int64
,
error
)
{
gpuInfo
:=
GetGPUInfo
()
if
gpuInfo
.
FreeMemory
>
0
&&
(
gpuInfo
.
Library
==
"cuda"
||
gpuInfo
.
Library
==
"rocm"
)
{
// leave 10% or
512
MiB of VRAM free per GPU to handle unaccounted for overhead
// leave 10% or
1024
MiB of VRAM free per GPU to handle unaccounted for overhead
overhead
:=
gpuInfo
.
FreeMemory
/
10
gpus
:=
uint64
(
gpuInfo
.
DeviceCount
)
if
overhead
<
gpus
*
512
*
1024
*
1024
{
overhead
=
gpus
*
512
*
1024
*
1024
if
overhead
<
gpus
*
1024
*
1024
*
1024
{
overhead
=
gpus
*
1024
*
1024
*
1024
}
return
int64
(
gpuInfo
.
FreeMemory
-
overhead
),
nil
}
...
...
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