Skip to content

updates to model for final experiments #6

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
vsoch merged 2 commits into from
Jul 23, 2025
Merged

updates to model for final experiments #6

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1766b98
updates to model for final experiments
vsoch Jul 23, 2025
625c45a
feat: add additional models.
vsoch Jul 23, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
45 changes: 45 additions & 0 deletions cmd/ocifit/main.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -51,6 +51,7 @@ const (

var (
universalDeserializer = serializer.NewCodecFactory(runtime.NewScheme()).UniversalDeserializer()
skipInstanceTypes = make(map[string]bool)
)

// WebhookServer with Node Cache and a direct k8s client
Expand Down Expand Up @@ -408,12 +409,29 @@ func (ws *WebhookServer) getAllNodeFeatures(ctx context.Context) ([]map[string]i
if err := ws.k8sClient.List(ctx, &nodeList, client.MatchingFields{"spec.unschedulable": "false"}); err != nil {
return nil, err
}

// We will do a filtering here for types we don't want to include
identityLabelKey := os.Getenv("NODE_IDENTITY_LABEL")
if identityLabelKey == "" {
identityLabelKey = instanceTypeLabel
}

var featureMatrix []map[string]interface{}
for _, node := range nodeList.Items {
features := make(map[string]interface{})
for key, val := range node.Labels {
features[key] = val
}

// Be conservative - don't include nodes we cannot identify
identityValue, ok := features[identityLabelKey].(string)
if !ok {
continue
}
if skipInstanceTypes[identityValue] {
log.Printf("Skipping dynamically discovered instance '%s' as it is in the skip list.", identityValue)
continue
}
featureMatrix = append(featureMatrix, features)
}
return featureMatrix, nil
Expand All @@ -422,6 +440,7 @@ func (ws *WebhookServer) getAllNodeFeatures(ctx context.Context) ([]map[string]i
// getCombinedNodeFeatures merges dynamically discovered nodes with a static catalog from a directory.
// It correctly handles static files that contain a JSON array (list) of node feature objects.
func (ws *WebhookServer) getCombinedNodeFeatures(ctx context.Context, staticFeaturesDir string) ([]map[string]interface{}, error) {

// Get all currently running schedulable nodes.
dynamicallyDiscoveredFeatures, err := ws.getAllNodeFeatures(ctx)
if err != nil {
Expand Down Expand Up @@ -484,6 +503,14 @@ func (ws *WebhookServer) getCombinedNodeFeatures(ctx context.Context, staticFeat
}

for _, staticFeatures := range featureList {
identityValue, ok := staticFeatures[identityLabelKey].(string)
if !ok {
continue
}
if skipInstanceTypes[identityValue] {
log.Printf("Skipping static instance '%s' from file '%s' as it is in the skip list.", identityValue, file.Name())
continue
}

// De-duplication check...
if identityValue, ok := staticFeatures[identityLabelKey].(string); ok {
Expand Down Expand Up @@ -606,7 +633,25 @@ func (ws *WebhookServer) handleMutate(w http.ResponseWriter, r *http.Request) {
w.Write(respBody)
}

func discoverSkipInstances() {
skipList := os.Getenv("SKIP_INSTANCE_TYPES")
if skipList != "" {
instances := strings.Split(skipList, ",")
for _, instance := range instances {
trimmed := strings.TrimSpace(instance)
if trimmed != "" {
log.Printf("Configuration: Will skip instance type '%s'", trimmed)
skipInstanceTypes[trimmed] = true
}
}
}
}

func main() {

// Ensure we parse skip instances
discoverSkipInstances()

config, err := clientcmd.BuildConfigFromFlags("", os.Getenv("KUBECONFIG"))
if err != nil {
log.Fatalf("Error building kubeconfig: %s", err.Error())
Expand Down
Binary file modified mlserver/models/lasso_model_fom.joblib
View file
Open in desktop
Binary file not shown.
Binary file modified mlserver/models/lasso_model_fom_per_dollar.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_gflops_per_second_ddot.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_gflops_per_second_mg.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_gflops_per_second_spmv.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_gflops_per_second_waxpby.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_memory_bandwidth_across_kernels_read.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_memory_bandwidth_across_kernels_total.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_memory_bandwidth_across_kernels_write.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_memory_used_data_total_gbytes.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_mpi_allreduce_avg.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_mpi_allreduce_max.joblib
View file
Open in desktop
Binary file not shown.
Binary file added mlserver/models/lasso_model_total_cg_iterations.joblib
View file
Open in desktop
Binary file not shown.