mirror of
https://github.com/mudler/LocalAI.git
synced 2026-06-12 18:58:49 -04:00
69e482b0a8
When a model config declares no explicit `backend:`, Load() fell into a
trial loop built by ranging the external-backends Go map (random order)
with no filtering, returning the first backend whose gRPC LoadModel
succeeded. An unrelated installed backend - e.g. the "opus" audio codec -
could therefore win a GGUF/LLM model load, so a model that should run on
llama.cpp wrongly tried to use opus.
Extract the candidate selection into a pure, testable function
SelectAutoLoadBackends that:
- sorts the candidate list deterministically (no more map-order
nondeterminism), and
- for a `.gguf` model, filters to LLM-capable backends (via
core/config.BackendCapabilities) and puts llama-cpp first, so an
incompatible audio/codec/image backend can never win the trial loop.
If filtering would leave zero candidates, the full sorted set is returned
unchanged, so a previously-loadable model is never made unloadable.
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
Assisted-by: claude:claude-opus-4-8 [Claude Code]
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
47 lines
1.8 KiB
Go
47 lines
1.8 KiB
Go
package model_test
|
|
|
|
import (
|
|
"github.com/mudler/LocalAI/pkg/model"
|
|
|
|
. "github.com/onsi/ginkgo/v2"
|
|
. "github.com/onsi/gomega"
|
|
)
|
|
|
|
var _ = Describe("SelectAutoLoadBackends (#9287)", func() {
|
|
Describe("GGUF model auto-detection", func() {
|
|
It("excludes incompatible audio/codec backends (e.g. opus) for a .gguf model", func() {
|
|
// Regression for #9287: installing an unrelated audio backend like
|
|
// "opus" must never win the GGUF auto-detect trial loop.
|
|
got := model.SelectAutoLoadBackends([]string{"opus", "llama-cpp"}, "Qwen3.5-9b.gguf")
|
|
Expect(got).NotTo(ContainElement("opus"))
|
|
Expect(got).To(ContainElement("llama-cpp"))
|
|
})
|
|
|
|
It("places llama-cpp first for a .gguf model", func() {
|
|
got := model.SelectAutoLoadBackends([]string{"vllm", "opus", "llama-cpp"}, "model.gguf")
|
|
Expect(got).NotTo(BeEmpty())
|
|
Expect(got[0]).To(Equal("llama-cpp"))
|
|
})
|
|
|
|
It("is deterministic regardless of input ordering", func() {
|
|
a := model.SelectAutoLoadBackends([]string{"opus", "vllm", "llama-cpp", "whisper"}, "m.gguf")
|
|
b := model.SelectAutoLoadBackends([]string{"whisper", "llama-cpp", "vllm", "opus"}, "m.gguf")
|
|
Expect(a).To(Equal(b))
|
|
})
|
|
|
|
It("falls back to the full sorted set when filtering leaves no candidate", func() {
|
|
// No LLM-capable backend installed: never make a previously-loadable
|
|
// model unloadable, return the original set (sorted).
|
|
got := model.SelectAutoLoadBackends([]string{"opus"}, "model.gguf")
|
|
Expect(got).To(Equal([]string{"opus"}))
|
|
})
|
|
})
|
|
|
|
Describe("non-GGUF model auto-detection", func() {
|
|
It("returns a deterministic (sorted) set without filtering", func() {
|
|
got := model.SelectAutoLoadBackends([]string{"opus", "llama-cpp", "diffusers"}, "model-dir")
|
|
Expect(got).To(Equal([]string{"diffusers", "llama-cpp", "opus"}))
|
|
})
|
|
})
|
|
})
|