dearwolf/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2025-05-20 02:24:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 21

chore(refactor): track grpcProcess in the model structure (#3663)

Browse source 
* chore(refactor): track grpcProcess in the model structure

This avoids to have to handle in two parts the data relative to the same
model. It makes it easier to track and use mutex with.

This also fixes races conditions while accessing to the model.

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* chore(tests): run protogen-go before starting aio tests

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* chore(tests): install protoc in aio tests

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

---------

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto 2024-09-26 12:44:55 +02:00 • committed by GitHub
parent 3d12d2037c
commit fa5c98549a
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
7 changed files with 71 additions and 44 deletions
Download patch file Download diff file Expand all files Collapse all files

15
pkg/model/initializers.go
View file

@ -304,18 +304,19 @@ func (ml *ModelLoader) grpcModel(backend string, o *Options) func(string, string
return nil, fmt.Errorf("failed allocating free ports: %s", err.Error())
}
// Make sure the process is executable
if err := ml.startProcess(uri, o.model, serverAddress); err != nil {
process, err := ml.startProcess(uri, o.model, serverAddress)
if err != nil {
log.Error().Err(err).Str("path", uri).Msg("failed to launch ")
return nil, err
}
log.Debug().Msgf("GRPC Service Started")
client = NewModel(modelName, serverAddress)
client = NewModel(modelName, serverAddress, process)
} else {
log.Debug().Msg("external backend is uri")
// address
client = NewModel(modelName, uri)
client = NewModel(modelName, uri, nil)
}
} else {
grpcProcess := backendPath(o.assetDir, backend)
@ -346,13 +347,14 @@ func (ml *ModelLoader) grpcModel(backend string, o *Options) func(string, string
args, grpcProcess = library.LoadLDSO(o.assetDir, args, grpcProcess)
// Make sure the process is executable in any circumstance
if err := ml.startProcess(grpcProcess, o.model, serverAddress, args...); err != nil {
process, err := ml.startProcess(grpcProcess, o.model, serverAddress, args...)
if err != nil {
return nil, err
}
log.Debug().Msgf("GRPC Service Started")
client = NewModel(modelName, serverAddress)
client = NewModel(modelName, serverAddress, process)
}
log.Debug().Msgf("Wait for the service to start up")
@ -374,6 +376,7 @@ func (ml *ModelLoader) grpcModel(backend string, o *Options) func(string, string
if !ready {
log.Debug().Msgf("GRPC Service NOT ready")
ml.deleteProcess(o.model)
return nil, fmt.Errorf("grpc service not ready")
}
@ -385,9 +388,11 @@ func (ml *ModelLoader) grpcModel(backend string, o *Options) func(string, string
res, err := client.GRPC(o.parallelRequests, ml.wd).LoadModel(o.context, &options)
if err != nil {
ml.deleteProcess(o.model)
return nil, fmt.Errorf("could not load model: %w", err)
}
if !res.Success {
ml.deleteProcess(o.model)
return nil, fmt.Errorf("could not load model (no success): %s", res.Message)
}