mirror of
https://github.com/dogkeeper886/ollama37.git
synced 2025-12-10 07:46:59 +00:00
config: allow setting context length through env var (#8938)
* envconfig: allow setting context length through env var
This commit is contained in:
@@ -167,6 +167,8 @@ var (
|
||||
MultiUserCache = Bool("OLLAMA_MULTIUSER_CACHE")
|
||||
// Enable the new Ollama engine
|
||||
NewEngine = Bool("OLLAMA_NEW_ENGINE")
|
||||
// ContextLength sets the default context length
|
||||
ContextLength = Uint("OLLAMA_CONTEXT_LENGTH", 2048)
|
||||
)
|
||||
|
||||
func String(s string) func() string {
|
||||
@@ -252,6 +254,7 @@ func AsMap() map[string]EnvVar {
|
||||
"OLLAMA_ORIGINS": {"OLLAMA_ORIGINS", AllowedOrigins(), "A comma separated list of allowed origins"},
|
||||
"OLLAMA_SCHED_SPREAD": {"OLLAMA_SCHED_SPREAD", SchedSpread(), "Always schedule model across all GPUs"},
|
||||
"OLLAMA_MULTIUSER_CACHE": {"OLLAMA_MULTIUSER_CACHE", MultiUserCache(), "Optimize prompt caching for multi-user scenarios"},
|
||||
"OLLAMA_CONTEXT_LENGTH": {"OLLAMA_CONTEXT_LENGTH", ContextLength(), "Context length to use unless otherwise specified (default: 2048)"},
|
||||
"OLLAMA_NEW_ENGINE": {"OLLAMA_NEW_ENGINE", NewEngine(), "Enable the new Ollama engine"},
|
||||
|
||||
// Informational
|
||||
|
||||
@@ -272,3 +272,19 @@ func TestVar(t *testing.T) {
|
||||
})
|
||||
}
|
||||
}
|
||||
|
||||
func TestContextLength(t *testing.T) {
|
||||
cases := map[string]uint{
|
||||
"": 2048,
|
||||
"4096": 4096,
|
||||
}
|
||||
|
||||
for k, v := range cases {
|
||||
t.Run(k, func(t *testing.T) {
|
||||
t.Setenv("OLLAMA_CONTEXT_LENGTH", k)
|
||||
if i := ContextLength(); i != v {
|
||||
t.Errorf("%s: expected %d, got %d", k, v, i)
|
||||
}
|
||||
})
|
||||
}
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user