mirror/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2025-12-30 22:20:20 -06:00
Code Issues Packages Projects Releases Wiki Activity

chore(importers/llama.cpp): add models to 'llama-cpp' subfolder (#7450)

Browse Source 
This makes paths predictable, and avoids multiple model files to show in
the main view

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto
2025-12-07 21:44:57 +01:00
committed by GitHub
parent 18b8956bd9
commit 8ca98c90ea
2 changed files with 11 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
core/gallery/importers/importers_test.go
View File

@@ -25,7 +25,7 @@ var _ = Describe("DiscoverModelConfig", func() {
Expect(modelConfig.Description).To(Equal("Imported from https://huggingface.co/mudler/LocalAI-functioncall-qwen2.5-7b-v0.5-Q4_K_M-GGUF"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("backend: llama-cpp"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(len(modelConfig.Files)).To(Equal(1), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].Filename).To(Equal("localai-functioncall-qwen2.5-7b-v0.5-q4_k_m.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].Filename).To(Equal("llama-cpp/models/localai-functioncall-qwen2.5-7b-v0.5-q4_k_m.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].URI).To(Equal("https://huggingface.co/mudler/LocalAI-functioncall-qwen2.5-7b-v0.5-Q4_K_M-GGUF/resolve/main/localai-functioncall-qwen2.5-7b-v0.5-q4_k_m.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].SHA256).To(Equal("4e7b7fe1d54b881f1ef90799219dc6cc285d29db24f559c8998d1addb35713d4"), fmt.Sprintf("Model config: %+v", modelConfig))
})
@@ -40,13 +40,13 @@ var _ = Describe("DiscoverModelConfig", func() {
Expect(modelConfig.Name).To(Equal("Qwen3-VL-2B-Instruct-GGUF"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Description).To(Equal("Imported from https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct-GGUF"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("backend: llama-cpp"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("mmproj: mmproj/mmproj-Qwen3VL-2B-Instruct-Q8_0.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("model: Qwen3VL-2B-Instruct-Q4_K_M.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("mmproj: llama-cpp/mmproj/mmproj-Qwen3VL-2B-Instruct-Q8_0.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("model: llama-cpp/models/Qwen3VL-2B-Instruct-Q4_K_M.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(len(modelConfig.Files)).To(Equal(2), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].Filename).To(Equal("Qwen3VL-2B-Instruct-Q4_K_M.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].Filename).To(Equal("llama-cpp/models/Qwen3VL-2B-Instruct-Q4_K_M.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].URI).To(Equal("https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct-GGUF/resolve/main/Qwen3VL-2B-Instruct-Q4_K_M.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].SHA256).ToNot(BeEmpty(), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[1].Filename).To(Equal("mmproj/mmproj-Qwen3VL-2B-Instruct-Q8_0.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[1].Filename).To(Equal("llama-cpp/mmproj/mmproj-Qwen3VL-2B-Instruct-Q8_0.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[1].URI).To(Equal("https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct-GGUF/resolve/main/mmproj-Qwen3VL-2B-Instruct-Q8_0.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[1].SHA256).ToNot(BeEmpty(), fmt.Sprintf("Model config: %+v", modelConfig))
})
@@ -61,13 +61,13 @@ var _ = Describe("DiscoverModelConfig", func() {
Expect(modelConfig.Name).To(Equal("Qwen3-VL-2B-Instruct-GGUF"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Description).To(Equal("Imported from https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct-GGUF"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("backend: llama-cpp"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("mmproj: mmproj/mmproj-Qwen3VL-2B-Instruct-F16.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("model: Qwen3VL-2B-Instruct-Q8_0.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("mmproj: llama-cpp/mmproj/mmproj-Qwen3VL-2B-Instruct-F16.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.ConfigFile).To(ContainSubstring("model: llama-cpp/models/Qwen3VL-2B-Instruct-Q8_0.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(len(modelConfig.Files)).To(Equal(2), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].Filename).To(Equal("Qwen3VL-2B-Instruct-Q8_0.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].Filename).To(Equal("llama-cpp/models/Qwen3VL-2B-Instruct-Q8_0.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].URI).To(Equal("https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct-GGUF/resolve/main/Qwen3VL-2B-Instruct-Q8_0.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[0].SHA256).ToNot(BeEmpty(), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[1].Filename).To(Equal("mmproj/mmproj-Qwen3VL-2B-Instruct-F16.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[1].Filename).To(Equal("llama-cpp/mmproj/mmproj-Qwen3VL-2B-Instruct-F16.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[1].URI).To(Equal("https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct-GGUF/resolve/main/mmproj-Qwen3VL-2B-Instruct-F16.gguf"), fmt.Sprintf("Model config: %+v", modelConfig))
Expect(modelConfig.Files[1].SHA256).ToNot(BeEmpty(), fmt.Sprintf("Model config: %+v", modelConfig))
})

4
core/gallery/importers/llama-cpp.go
View File

@@ -184,7 +184,7 @@ func (i *LlamaCPPImporter) Import(details Details) (gallery.ModelConfig, error)
if strings.Contains(strings.ToLower(file.Path), "mmproj") {
lastMMProjFile = &gallery.File{
URI: file.URL,
Filename: filepath.Join("mmproj", filepath.Base(file.Path)),
Filename: filepath.Join("llama-cpp", "mmproj", filepath.Base(file.Path)),
SHA256: file.SHA256,
}
if slices.ContainsFunc(mmprojQuantsList, func(quant string) bool {
@@ -196,7 +196,7 @@ func (i *LlamaCPPImporter) Import(details Details) (gallery.ModelConfig, error)
} else if strings.HasSuffix(strings.ToLower(file.Path), "gguf") {
lastGGUFFile = &gallery.File{
URI: file.URL,
Filename: filepath.Base(file.Path),
Filename: filepath.Join("llama-cpp", "models", filepath.Base(file.Path)),
SHA256: file.SHA256,
}
// get the files of the prefered quants