baalajimaestro/ollama
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

tests: fix max queue integration test (#7782)

Browse source 
This had fallen out of sync with the envconfig behavior, where max queue default was not zero.
This commit is contained in:
Daniel Hiltgen 2024-11-22 08:05:45 -08:00 committed by GitHub
parent b85520bfb9
commit f0a351810c
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 2 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
integration/max_queue_test.go
View file

@ -16,7 +16,6 @@ import (
"github.com/stretchr/testify/require" "github.com/stretchr/testify/require"
"github.com/ollama/ollama/api" "github.com/ollama/ollama/api"
"github.com/ollama/ollama/envconfig"
) )
func TestMaxQueue(t *testing.T) { func TestMaxQueue(t *testing.T) {
@ -27,12 +26,8 @@ func TestMaxQueue(t *testing.T) {
// Note: This test can be quite slow when running in CPU mode, so keep the threadCount low unless your on GPU // Note: This test can be quite slow when running in CPU mode, so keep the threadCount low unless your on GPU
// Also note that by default Darwin can't sustain > ~128 connections without adjusting limits // Also note that by default Darwin can't sustain > ~128 connections without adjusting limits
threadCount := 32 threadCount := 16
if maxQueue := envconfig.MaxQueue(); maxQueue != 0 { t.Setenv("OLLAMA_MAX_QUEUE", strconv.Itoa(threadCount))
threadCount = int(maxQueue)
} else {
t.Setenv("OLLAMA_MAX_QUEUE", strconv.Itoa(threadCount))
}
req := api.GenerateRequest{ req := api.GenerateRequest{
Model: "orca-mini", Model: "orca-mini",